Add FUNDING.yml

Resorted to reinstalling `torchvision` via `pip`

.deepsource.toml

@@ -0,0 +1,12 @@
+version = 1
+
+[[analyzers]]
+name = "python"
+
+  [analyzers.meta]
+  runtime_version = "3.x.x"
+  max_line_length = 135
+
+
+[[analyzers]]
+name = "docker"

.devcontainer/.devcontainer.json

@@ -0,0 +1,45 @@
+// For format details, see https://aka.ms/devcontainer.json. For config options, see the
+// README at: https://github.com/devcontainers/templates/tree/main/src/python
+{
+	"name": "Python 3",
+	// Or use a Dockerfile or Docker Compose file. More info: https://containers.dev/guide/dockerfile
+	"image": "mcr.microsoft.com/devcontainers/python:0-3.11",
+	"features": {
+		"ghcr.io/devcontainers-contrib/features/poetry:2": {
+			"version": "latest"
+		},
+		"ghcr.io/devcontainers-contrib/features/nox:2": {
+			"version": "latest"
+		},
+    	"ghcr.io/devcontainers/features/github-cli:1": {
+			"version": "latest"
+		}
+	},
+
+	// Features to add to the dev container. More info: https://containers.dev/features.
+	// "features": {},
+
+	// Use 'forwardPorts' to make a list of ports inside the container available locally.
+	// "forwardPorts": [],
+
+	// Use 'postCreateCommand' to run commands after the container is created.
+	"postCreateCommand": "poetry install",
+
+	// Configure tool-specific properties.
+	"customizations": {
+		"vscode": {
+			"settings": {},
+			"extensions": [
+				"ms-python.python",
+				"ms-python.vscode-pylance"
+			]
+    	}
+  	},
+	"mounts": [
+		// Re-use local Git configuration
+		"source=${localEnv:HOME}/.gitconfig,target=/home/vscode/.gitconfig,type=bind,consistency=cached"
+	]
+
+	// Uncomment to connect as root instead. More info: https://aka.ms/dev-containers-non-root.
+	// "remoteUser": "vscode"
+}

.dockerignore

@@ -1,3 +1,6 @@
 Dockerfile
 .venv
 docker-compose.yml
+*.ipnyb
+dist/
+*.pkl

.github/FUNDING.yml

@@ -0,0 +1 @@
+github: [slashtechno]

.github/workflows/python-publish.yml

@@ -8,12 +8,15 @@
 
 name: Upload Python Package
 
+
 on:
   release:
     types: [published]
+  workflow_dispatch:
 
 permissions:
   contents: read
+  id-token: write
 
 jobs:
   deploy:

.vscode/launch.json

@@ -4,14 +4,44 @@
     // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
     "version": "0.2.0",
     "configurations": [
+        {
+            "name": "Quick Debug",
+            "type": "python",
+            "request": "launch",
+            "module": "wyzely_detect",
+            "args": [
+                "--run-scale", "0.25", "--view-scale", "0.5", "--no-remove-representations", "--fake-second-source"
+            ],
+            "justMyCode": true
+        },
+        // {
+        //     "name": "Quick, Specific Debug",
+        //     "type": "python",
+        //     "request": "launch",
+        //     "module": "wyzely_detect",
+        //     "args": [
+        //         "--run-scale", "0.25", "--view-scale", "0.5", "--no-remove-representations", "--detect-object", "person", "--detect-object", "cell phone"
+        //     ],
+        //     "justMyCode": true
+        // },
         {
             // "name": "Python: Module",
-            "name": "Debug Wyzely Detect",
+            "name": "Full Debug",
             "type": "python",
             "request": "launch",
             "module": "wyzely_detect",
             // "justMyCode": true
             "justMyCode": false
-        }
+        },
+        {
+            "name": "Debug --help",
+            "type": "python",
+            "request": "launch",
+            "module": "wyzely_detect",
+            "args": [
+                "--help"
+            ],
+            "justMyCode": false
+        },
     ]
 }

Dockerfile

@@ -1,12 +1,19 @@
 FROM python:3.10.5-buster
 
+LABEL org.opencontainers.image.description "Docker image for running wyzely-detect"
+LABEL org.opencontainers.image.source "https://github.com/slashtechno/wyzely-detect"
+
 RUN apt update && apt install libgl1 -y
 RUN pip install poetry
 
+
 WORKDIR /app
 
 COPY . .
 
 RUN poetry install
 
-ENTRYPOINT ["poetry", "run", "python", "-m", "wyzely_detect"]
+# RUN poetry run pip uninstall -y torchvision 
+# RUN poetry run pip install torchvision 
+
+ENTRYPOINT ["poetry", "run", "python", "-m", "--", "wyzely_detect", "--no-display"]

LICENSE

@@ -1,5 +1,5 @@
-                    GNU GENERAL PUBLIC LICENSE
+                    GNU AFFERO GENERAL PUBLIC LICENSE
-                       Version 3, 29 June 2007
+                       Version 3, 19 November 2007
 
  Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>
  Everyone is permitted to copy and distribute verbatim copies
@@ -7,17 +7,15 @@
 
                             Preamble
 
-  The GNU General Public License is a free, copyleft license for
+  The GNU Affero General Public License is a free, copyleft license for
-software and other kinds of works.
+software and other kinds of works, specifically designed to ensure
+cooperation with the community in the case of network server software.
 
   The licenses for most software and other practical works are designed
 to take away your freedom to share and change the works.  By contrast,
-the GNU General Public License is intended to guarantee your freedom to
+our General Public Licenses are intended to guarantee your freedom to
 share and change all versions of a program--to make sure it remains free
-software for all its users.  We, the Free Software Foundation, use the
+software for all its users.
-GNU General Public License for most of our software; it applies also to
-any other work released this way by its authors.  You can apply it to
-your programs, too.
 
   When we speak of free software, we are referring to freedom, not
 price.  Our General Public Licenses are designed to make sure that you
@@ -26,44 +24,34 @@ them if you wish), that you receive source code or can get it if you
 want it, that you can change the software or use pieces of it in new
 free programs, and that you know you can do these things.
 
-  To protect your rights, we need to prevent others from denying you
+  Developers that use our General Public Licenses protect your rights
-these rights or asking you to surrender the rights.  Therefore, you have
+with two steps: (1) assert copyright on the software, and (2) offer
-certain responsibilities if you distribute copies of the software, or if
+you this License which gives you legal permission to copy, distribute
-you modify it: responsibilities to respect the freedom of others.
+and/or modify the software.
 
-  For example, if you distribute copies of such a program, whether
+  A secondary benefit of defending all users' freedom is that
-gratis or for a fee, you must pass on to the recipients the same
+improvements made in alternate versions of the program, if they
-freedoms that you received.  You must make sure that they, too, receive
+receive widespread use, become available for other developers to
-or can get the source code.  And you must show them these terms so they
+incorporate.  Many developers of free software are heartened and
-know their rights.
+encouraged by the resulting cooperation.  However, in the case of
+software used on network servers, this result may fail to come about.
+The GNU General Public License permits making a modified version and
+letting the public access it on a server without ever releasing its
+source code to the public.
 
-  Developers that use the GNU GPL protect your rights with two steps:
+  The GNU Affero General Public License is designed specifically to
-(1) assert copyright on the software, and (2) offer you this License
+ensure that, in such cases, the modified source code becomes available
-giving you legal permission to copy, distribute and/or modify it.
+to the community.  It requires the operator of a network server to
+provide the source code of the modified version running there to the
+users of that server.  Therefore, public use of a modified version, on
+a publicly accessible server, gives the public access to the source
+code of the modified version.
 
-  For the developers' and authors' protection, the GPL clearly explains
+  An older license, called the Affero General Public License and
-that there is no warranty for this free software.  For both users' and
+published by Affero, was designed to accomplish similar goals.  This is
-authors' sake, the GPL requires that modified versions be marked as
+a different license, not a version of the Affero GPL, but Affero has
-changed, so that their problems will not be attributed erroneously to
+released a new version of the Affero GPL which permits relicensing under
-authors of previous versions.
+this license.
-
-  Some devices are designed to deny users access to install or run
-modified versions of the software inside them, although the manufacturer
-can do so.  This is fundamentally incompatible with the aim of
-protecting users' freedom to change the software.  The systematic
-pattern of such abuse occurs in the area of products for individuals to
-use, which is precisely where it is most unacceptable.  Therefore, we
-have designed this version of the GPL to prohibit the practice for those
-products.  If such problems arise substantially in other domains, we
-stand ready to extend this provision to those domains in future versions
-of the GPL, as needed to protect the freedom of users.
-
-  Finally, every program is threatened constantly by software patents.
-States should not allow patents to restrict development and use of
-software on general-purpose computers, but in those that do, we wish to
-avoid the special danger that patents applied to a free program could
-make it effectively proprietary.  To prevent this, the GPL assures that
-patents cannot be used to render the program non-free.
 
   The precise terms and conditions for copying, distribution and
 modification follow.
@@ -72,7 +60,7 @@ modification follow.
 
   0. Definitions.
 
-  "This License" refers to version 3 of the GNU General Public License.
+  "This License" refers to version 3 of the GNU Affero General Public License.
 
   "Copyright" also means copyright-like laws that apply to other kinds of
 works, such as semiconductor masks.
@@ -549,35 +537,45 @@ to collect a royalty for further conveying from those to whom you convey
 the Program, the only way you could satisfy both those terms and this
 License would be to refrain entirely from conveying the Program.
 
-  13. Use with the GNU Affero General Public License.
+  13. Remote Network Interaction; Use with the GNU General Public License.
+
+  Notwithstanding any other provision of this License, if you modify the
+Program, your modified version must prominently offer all users
+interacting with it remotely through a computer network (if your version
+supports such interaction) an opportunity to receive the Corresponding
+Source of your version by providing access to the Corresponding Source
+from a network server at no charge, through some standard or customary
+means of facilitating copying of software.  This Corresponding Source
+shall include the Corresponding Source for any work covered by version 3
+of the GNU General Public License that is incorporated pursuant to the
+following paragraph.
 
   Notwithstanding any other provision of this License, you have
 permission to link or combine any covered work with a work licensed
-under version 3 of the GNU Affero General Public License into a single
+under version 3 of the GNU General Public License into a single
 combined work, and to convey the resulting work.  The terms of this
 License will continue to apply to the part which is the covered work,
-but the special requirements of the GNU Affero General Public License,
+but the work with which it is combined will remain governed by version
-section 13, concerning interaction through a network will apply to the
+3 of the GNU General Public License.
-combination as such.
 
   14. Revised Versions of this License.
 
   The Free Software Foundation may publish revised and/or new versions of
-the GNU General Public License from time to time.  Such new versions will
+the GNU Affero General Public License from time to time.  Such new versions
-be similar in spirit to the present version, but may differ in detail to
+will be similar in spirit to the present version, but may differ in detail to
 address new problems or concerns.
 
   Each version is given a distinguishing version number.  If the
-Program specifies that a certain numbered version of the GNU General
+Program specifies that a certain numbered version of the GNU Affero General
 Public License "or any later version" applies to it, you have the
 option of following the terms and conditions either of that numbered
 version or of any later version published by the Free Software
 Foundation.  If the Program does not specify a version number of the
-GNU General Public License, you may choose any version ever published
+GNU Affero General Public License, you may choose any version ever published
 by the Free Software Foundation.
 
   If the Program specifies that a proxy can decide which future
-versions of the GNU General Public License can be used, that proxy's
+versions of the GNU Affero General Public License can be used, that proxy's
 public statement of acceptance of a version permanently authorizes you
 to choose that version for the Program.
 
@@ -635,40 +633,29 @@ the "copyright" line and a pointer to where the full notice is found.
     Copyright (C) <year>  <name of author>
 
     This program is free software: you can redistribute it and/or modify
-    it under the terms of the GNU General Public License as published by
+    it under the terms of the GNU Affero General Public License as published
-    the Free Software Foundation, either version 3 of the License, or
+    by the Free Software Foundation, either version 3 of the License, or
     (at your option) any later version.
 
     This program is distributed in the hope that it will be useful,
     but WITHOUT ANY WARRANTY; without even the implied warranty of
     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-    GNU General Public License for more details.
+    GNU Affero General Public License for more details.
 
-    You should have received a copy of the GNU General Public License
+    You should have received a copy of the GNU Affero General Public License
     along with this program.  If not, see <https://www.gnu.org/licenses/>.
 
 Also add information on how to contact you by electronic and paper mail.
 
-  If the program does terminal interaction, make it output a short
+  If your software can interact with users remotely through a computer
-notice like this when it starts in an interactive mode:
+network, you should also make sure that it provides a way for users to
-
+get its source.  For example, if your program is a web application, its
-    <program>  Copyright (C) <year>  <name of author>
+interface could display a "Source" link that leads users to an archive
-    This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
+of the code.  There are many ways you could offer source, and different
-    This is free software, and you are welcome to redistribute it
+solutions will be better for different programs; see section 13 for the
-    under certain conditions; type `show c' for details.
+specific requirements.
-
-The hypothetical commands `show w' and `show c' should show the appropriate
-parts of the General Public License.  Of course, your program's commands
-might be different; for a GUI interface, you would use an "about box".
 
   You should also get your employer (if you work as a programmer) or school,
 if any, to sign a "copyright disclaimer" for the program, if necessary.
-For more information on this, and how to apply and follow the GNU GPL, see
+For more information on this, and how to apply and follow the GNU AGPL, see
 <https://www.gnu.org/licenses/>.
-
-  The GNU General Public License does not permit incorporating your program
-into proprietary programs.  If your program is a subroutine library, you
-may consider it more useful to permit linking proprietary applications with
-the library.  If this is what you want to do, use the GNU Lesser General
-Public License instead of this License.  But first, please read
-<https://www.gnu.org/licenses/why-not-lgpl.html>.

README.md

@@ -11,11 +11,16 @@ Recognize faces/objects in a video stream (from a webcam or a security camera) a
 
 
 ## Prerequisites  
-### Poetry/Python  
+### Python  
 - Camera, either a webcam or a Wyze Cam  
     - All RTSP feeds _should_ work, however.  
+    - **WSL, by default, does not support USB devices.** It is recommended to natively run this, but it is possible to use it on WSL with streams or some workarounds.  
 - Python 3.10 or 3.11  
-- Poetry  
+- Poetry (optional)  
+- Windows or Linux  
+    - I've tested this on MacOS - it works on my 2014 MacBook Air but not a 2011 MacBook Pro  
+    - Both were upgraded with OpenCore, with the MacBook Air running Monterey and the MacBook Pro running a newer version of MacOS, which may have been the problem  
+
 ### Docker  
 - A Wyze Cam  
     - Any other RTSP feed _should_ work, as mentioned above  
@@ -28,17 +33,29 @@ Recognize faces/objects in a video stream (from a webcam or a security camera) a
 
 ## Usage  
 ### Installation  
+Cloning the repository is not required when installing from PyPi but is required when installing from source  
 1. Clone this repo with `git clone https://github.com/slashtechno/wyzely-detect`  
 2. `cd` into the cloned repository  
 3. Then, either install with [Poetry](https://python-poetry.org/) or run with Docker  
 
-#### Docker  
-1. Modify to `docker-compose.yml` to achieve desired configuration
-2. Run in the background with `docker compose up -d
 
-#### Poetry  
+#### Installing from PyPi with pip (recommended)  
+This assumes you have Python 3.10 or 3.11 installed  
+1. `pip install wyzely-detect`  
+    a. You may need to use `pip3` instead of `pip`  
+2. `wyzely-detect`  
+
+#### Poetry (best for GPU support)
 1. `poetry install`  
+    a. For GPU support, use `poetry install -E cuda --with gpu`
 2. `poetry run -- wyzely-detect`  
+
+#### Docker  
+Running with Docker has the benefit of having easier configuration, the ability to run headlessly, and easy setup of Ntfy and [mrlt8/docker-wyze-bridge](https://github.com/mrlt8/docker-wyze-bridge). However, for now, CPU-only is supported. Contributions are welcome to add GPU support. In addition, Docker is tested a less-tested method of running this program.  
+
+1. Modify to `docker-compose.yml` to achieve desired configuration  
+2. Run in the background with `docker compose up -d`
+
 ### Configuration  
 The following are some basic CLI options. Most flags have environment variable equivalents which can be helpful when using Docker. 
 

deepface-test.ipynb

@@ -36,7 +36,7 @@
     "# cv2.imwrite(str(uuid_path), frame)\n",
     "# dfs = DeepFace.find(img_path=str(uuid_path), db_path = \"faces\")\n",
     "# Don't throw an error if no face is detected (enforce_detection=False)\n",
-    "dfs = DeepFace.find(frame, db_path = \"faces\", enforce_detection=False, silent=False, model_name=\"ArcFace\", detector_backend=\"opencv\")\n",
+    "dfs = DeepFace.find(frame, db_path = \"faces\", enforce_detection=True, silent=False, model_name=\"ArcFace\", detector_backend=\"opencv\")\n",
     "# Get the identity of the person\n",
     "for i, pd_dataframe in enumerate(dfs):\n",
     "    # Sort the dataframe by confidence\n",

docker-compose.yml

@@ -6,16 +6,19 @@ services:
         container_name: bridge-wyzely-detect
         restart: unless-stopped
         image: mrlt8/wyze-bridge:latest
-        # I think we can remove the ports, since we're using the network
+        # The ports can be removed since we're using the network
-        # Just an unnecesary security risk
+        # Just an unnecesary security risk to expose them but can be useful for debugging
         # ports:
         #     - 1935:1935 # RTMP
         #     - 8554:8554 # RTSP (this is really the only one we need)
         #     - 8888:8888 # HLS
         # - 5000:5000 # WEB-UI
         environment:
-            - WYZE_EMAIL=${WYZE_EMAIL} # Replace with wyze email 
+            # This is a simple configuration without 2FA.
-            - WYZE_PASSWORD=${WYZE_PASSWORD} # Replace with wyze password
+            # For advanced configuration, including using an API key, see https://github.com/mrlt8/docker-wyze-bridge/wiki/Two-Factor-Authentication
+            # Either replace the following with your Wyze username and password, or set the environment variables
+            - WYZE_EMAIL=${WYZE_EMAIL} 
+            - WYZE_PASSWORD=${WYZE_PASSWORD}
         networks:
             all:
     ntfy:
@@ -36,18 +39,27 @@ services:
     wyzely-detect:
       container_name: wyzely-detect
       restart: unless-stopped 
-      # image: ghcr.io/slashtechno/wyzely-detect:latest
+      image: ghcr.io/slashtechno/wyzely-detect:latest
-      build:
+      # Building from source is also an option
-        context: .
+      # build:
-        dockerfile: Dockerfile
+      #   context: .
+      #   dockerfile: Dockerfile
+      command:
+        - "--ntfy-url"
+        # Replace "wyzely-detect" with the desired notification stream
+        - "http://ntfy:80/wyzely-detect"
+
+        - "--rtsp-url"
+        # Replace "cv" with the desired rtsp stream
+        - "rtsp://bridge:8554/cv"
+
+        # Example second rtsp stream
+        # - "--rtsp-url"
+        # - "rtsp://bridge:8554/camera"
       volumes:
         - ./faces:/app/faces
       networks:
         all:
-      environment:
-        - URL=rtsp://bridge:8554/cv
-        - NO_DISPLAY=true
-        - NTFY_URL=http://ntfy:80/wyzely-detect
       depends_on:
         - bridge
 

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "wyzely-detect"
-version = "0.1.0"
+version = "0.2.1"
 description = "Recognize faces/objects in a video stream (from a webcam or a security camera) and send notifications to your devices"
 authors = ["slashtechno <77907286+slashtechno@users.noreply.github.com>"]
 repository = "https://github.com/slashtechno/wyzely-detect"
@@ -21,17 +21,56 @@ ultralytics = "^8.0.190"
 hjson = "^3.1.0"
 numpy = "^1.23.2"
 
-# https://github.com/python-poetry/poetry/issues/6409
+# https://github.com/python-poetry/poetry/issues/6409#issuecomment-1911735833
-torch = ">=2.0.0, !=2.0.1, !=2.1.0"
+# If GPU support doesn't work, `poetry install -E cuda --with gpu` will force it to be installed from the GPU PyTorch repo
-
+# However, PyPi's `torch` has CUDA 12.1 support by default on Linux, so in that case it should not be needed.
+torch = [
+    {version = "^2.2.1", source = "pypi", markers = "extra!='cuda' and (platform_system=='Linux' or platform_system=='Darwin')"},
+    {version = "^2.2.1", source = "pytorch-cpu", markers = "extra!='cuda' and platform_system=='Windows'"},
+    ]
 # https://stackoverflow.com/a/76477590/18270659
-# https://discuss.tensorflow.org/t/tensorflow-io-gcs-filesystem-with-windows/18849/4
+# https://discfuss.tensorflow.org/t/tensorflow-io-gcs-filesystem-with-windows/18849/4
+# https://github.com/python-poetry/poetry/issues/8271#issuecomment-1712020965
 # Might be able to remove this version constraint later
-tensorflow-io-gcs-filesystem = "0.31.0"
+# Working versions:
-tensorflow = "^2.14.0"
+# Python version 3.10.12 and 3.10.5 both work
+# CUDA version - 12.2
+# cuDNN version - 8.8.1
+# Installed from Nvidia website - nvidia-cuda-toolkit is not installed, but default PopOS drivers are installed
+absl-py = "^2.1.0"
+tensorflow = {version = "^2.13.0", markers = "extra!='cuda'"}
+# TODO: Change platform to markers
+tensorflow-macos = { version = "^2.13.0", platform = "darwin", markers = "platform_machine=='arm64'" }
+tensorflow-intel = { version = "^2.13.0", platform = "win32" }
+tensorflow-io-gcs-filesystem = [
+    { version = "< 0.32.0", markers = "platform_system == 'Windows'" }
+]
 
 
 deepface = "^0.0.79"
+prettytable = "^3.9.0"
+
+
+[tool.poetry.group.gpu]
+optional = true
+
+[tool.poetry.group.gpu.dependencies]
+torch = {version = "^2.2.1", source = "pytorch-cu121", markers = "extra=='cuda'"}
+tensorflow = {version = "^2.14.0", extras = ["and-cuda"], markers = "extra=='cuda'  and platform_system == 'Linux'"}
+
+[tool.poetry.extras]
+# Might be better to rename this to nocpu since it's more accurate
+cuda = []
+
+[[tool.poetry.source]]
+name = "pytorch-cpu"
+url = "https://download.pytorch.org/whl/cpu"
+priority = "explicit"
+
+[[tool.poetry.source]]
+name = "pytorch-cu121"
+url = "https://download.pytorch.org/whl/cu121"
+priority = "explicit"
 
 [tool.poetry.group.dev.dependencies]
 black = "^23.9.1"

wyzely_detect/__main__.py

@@ -1,162 +1,22 @@
 # import face_recognition
-import cv2
-import dotenv
 from pathlib import Path
-import os
+import cv2
+import sys
+from prettytable import PrettyTable
 
 # import hjson as json
 import torch
 from ultralytics import YOLO
 
-import argparse
-
-from .utils import notify
 from .utils import utils
+from .utils.cli_args import argparser
 
 DATETIME_FORMAT = "%Y-%m-%d %H:%M:%S"
 args = None
 
-objects_and_peoples = {
-    "objects": {},
-    "peoples": {},
-}
-
 
 def main():
-    global objects_and_peoples
     global args
-    # RUN_BY_COMPOSE = os.getenv("RUN_BY_COMPOSE") # Replace this with code to check for gpu
-
-    if Path(".env").is_file():
-        dotenv.load_dotenv()
-        print("Loaded .env file")
-    else:
-        print("No .env file found")
-
-    # TODO: If possible, move the argparse stuff to a separate file
-    # It's taking up too many lines in this file
-    argparser = argparse.ArgumentParser(
-        prog="Wyzely Detect",
-        description="Recognize faces/objects in a video stream (from a webcam or a security camera) and send notifications to your devices",  # noqa: E501
-        epilog=":)",
-    )
-
-    # required='RUN_SCALE' not in os.environ,
-
-    argparser.add_argument(
-        "--run-scale",
-        # Set it to the env RUN_SCALE if it isn't blank, otherwise set it to 0.25
-        default=os.environ["RUN_SCALE"]
-        if "RUN_SCALE" in os.environ and os.environ["RUN_SCALE"] != ""
-        # else 0.25,
-        else 1,
-        type=float,
-        help="The scale to run the detection at, default is 0.25",
-    )
-    argparser.add_argument(
-        "--view-scale",
-        # Set it to the env VIEW_SCALE if it isn't blank, otherwise set it to 0.75
-        default=os.environ["VIEW_SCALE"]
-        if "VIEW_SCALE" in os.environ and os.environ["VIEW_SCALE"] != ""
-        # else 0.75,
-        else 1,
-        type=float,
-        help="The scale to view the detection at, default is 0.75",
-    )
-
-    argparser.add_argument(
-        "--no-display",
-        default=os.environ["NO_DISPLAY"]
-        if "NO_DISPLAY" in os.environ and os.environ["NO_DISPLAY"] != ""
-        else False,
-        action="store_true",
-        help="Don't display the video feed",
-    )
-
-    argparser.add_argument(
-        "--confidence-threshold",
-        default=os.environ["CONFIDENCE_THRESHOLD"]
-        if "CONFIDENCE_THRESHOLD" in os.environ
-        and os.environ["CONFIDENCE_THRESHOLD"] != ""
-        else 0.6,
-        type=float,
-        help="The confidence threshold to use",
-    )
-
-    argparser.add_argument(
-        "--faces-directory",
-        default=os.environ["FACES_DIRECTORY"]
-        if "FACES_DIRECTORY" in os.environ and os.environ["FACES_DIRECTORY"] != ""
-        else "faces",
-        type=str,
-        help="The directory to store the faces. Can either contain images or subdirectories with images, the latter being the preferred method",  # noqa: E501
-    )
-    argparser.add_argument(
-        "--detect-object",
-        nargs="*",
-        default=[],
-        type=str,
-        help="The object(s) to detect. Must be something the model is trained to detect",
-    )
-
-    stream_source = argparser.add_mutually_exclusive_group()
-    stream_source.add_argument(
-        "--url",
-        default=os.environ["URL"]
-        if "URL" in os.environ and os.environ["URL"] != ""
-        else None,  # noqa: E501
-        type=str,
-        help="The URL of the stream to use",
-    )
-    stream_source.add_argument(
-        "--capture-device",
-        default=os.environ["CAPTURE_DEVICE"]
-        if "CAPTURE_DEVICE" in os.environ and os.environ["CAPTURE_DEVICE"] != ""
-        else 0,  # noqa: E501
-        type=int,
-        help="The capture device to use. Can also be a url.",
-    )
-
-    # Defaults for the stuff here and down are already set in notify.py.
-    # Setting them here just means that argparse will display the default values as defualt
-    # TODO: Perhaps just remove the default parameter and just add to the help message that the default is set is x
-    # TODO: Make ntfy optional in ntfy.py. Currently, unless there is a local or LAN instance of ntfy, this can't run offline
-    notifcation_services = argparser.add_argument_group("Notification Services")
-    notifcation_services.add_argument(
-        "--ntfy-url",
-        default=os.environ["NTFY_URL"]
-        if "NTFY_URL" in os.environ and os.environ["NTFY_URL"] != ""
-        else "https://ntfy.sh/wyzely-detect",
-        type=str,
-        help="The URL to send notifications to",
-    )
-
-    timers = argparser.add_argument_group("Timers")
-    timers.add_argument(
-        "--detection-duration",
-        default=os.environ["DETECTION_DURATION"]
-        if "DETECTION_DURATION" in os.environ and os.environ["DETECTION_DURATION"] != ""
-        else 2,
-        type=int,
-        help="The duration (in seconds) that an object must be detected for before sending a notification",
-    )
-    timers.add_argument(
-        "--detection-window",
-        default=os.environ["DETECTION_WINDOW"]
-        if "DETECTION_WINDOW" in os.environ and os.environ["DETECTION_WINDOW"] != ""
-        else 15,
-        type=int,
-        help="The time (seconds) before the detection duration resets",
-    )
-    timers.add_argument(
-        "--notification-window",
-        default=os.environ["NOTIFICATION_WINDOW"]
-        if "NOTIFICATION_WINDOW" in os.environ
-        and os.environ["NOTIFICATION_WINDOW"] != ""
-        else 30,
-        type=int,
-        help="The time (seconds) before another notification can be sent",
-    )
 
     args = argparser.parse_args()
 
@@ -164,149 +24,111 @@ def main():
     # https://github.com/ultralytics/ultralytics/issues/3084#issuecomment-1732433168
     # Currently, I have been unable to set up Poetry to use GPU for Torch
     for i in range(torch.cuda.device_count()):
-        print(torch.cuda.get_device_properties(i).name)
+        print(f"Using {torch.cuda.get_device_properties(i).name} for pytorch")
     if torch.cuda.is_available():
         torch.cuda.set_device(0)
         print("Set CUDA device")
     else:
         print("No CUDA device available, using CPU")
+    # Seems automatically, deepface (tensorflow) tried to use my GPU on Pop!_OS (I did not set up cudnn or anything)
+    # Not sure the best way, in Poetry, to manage GPU libraries so for now, just use CPU
+    if args.force_disable_tensorflow_gpu:
+        print("Forcing tensorflow to use CPU")
+        import tensorflow as tf
+
+        tf.config.set_visible_devices([], "GPU")
+        if tf.config.experimental.list_logical_devices("GPU"):
+            print("GPU disabled unsuccessfully")
+        else:
+            print("GPU disabled successfully")
 
     model = YOLO("yolov8n.pt")
 
     # Depending on if the user wants to use a stream or a capture device,
     # Set the video capture to the appropriate source
-    if args.url:
+    if not args.rtsp_url and not args.capture_device:
-        video_capture = cv2.VideoCapture(args.url)
+        print("No stream or capture device set, defaulting to capture device 0")
+        video_sources = {"devices": [cv2.VideoCapture(0)]}
     else:
-        video_capture = cv2.VideoCapture(args.capture_device)
+        video_sources = {
+            "streams": [cv2.VideoCapture(url) for url in args.rtsp_url],
+            "devices": [cv2.VideoCapture(device) for device in args.capture_device],
+        }
+
+    if args.fake_second_source:
+        try:
+            video_sources["devices"].append(video_sources["devices"][0])
+        except KeyError:
+            print("No capture device to use as second source. Trying stream.")
+            try:
+                video_sources["devices"].append(video_sources["devices"][0])
+            except KeyError:
+                print("No stream to use as a second source")
+                # When the code tries to resize the nonexistent capture device 1, the program will fail
 
     # Eliminate lag by setting the buffer size to 1
     # This makes it so that the video capture will only grab the most recent frame
     # However, this means that the video may be choppy
-    video_capture.set(cv2.CAP_PROP_BUFFERSIZE, 1)
+    # Only do this for streams
+    try:
+        for stream in video_sources["streams"]:
+            stream.set(cv2.CAP_PROP_BUFFERSIZE, 1)
+    # If there are no streams, this will throw a KeyError
+    except KeyError:
+        pass
 
-    # Print the resolution of the video
+    # Print out the resolution of the video sources. Ideally, change this so the device ID/url is also printed
-    print(
+    pretty_table = PrettyTable(field_names=["Source Type", "Resolution"])
-        f"Video resolution: {video_capture.get(cv2.CAP_PROP_FRAME_WIDTH)}x{video_capture.get(cv2.CAP_PROP_FRAME_HEIGHT)}"  # noqa: E501
+    for source_type, sources in video_sources.items():
+        for source in sources:
+            if (
+                source.get(cv2.CAP_PROP_FRAME_WIDTH) == 0
+                or source.get(cv2.CAP_PROP_FRAME_HEIGHT) == 0
+            ):
+                message = "Capture for a source failed as resolution is 0x0.\n"
+                if source_type == "streams":
+                    message += "Check if the stream URL is correct and if the stream is online."
+                else:
+                    message += "Check if the capture device is connected, working, and not in use by another program."
+                print(message)
+                sys.exit(1)
+            pretty_table.add_row(
+                [
+                    source_type,
+                    f"{source.get(cv2.CAP_PROP_FRAME_WIDTH)}x{source.get(cv2.CAP_PROP_FRAME_HEIGHT)}",
+                ]
             )
-
+    print(pretty_table)
     print("Beginning video capture...")
     while True:
         # Grab a single frame of video
-        ret, frame = video_capture.read()
+        frames = []
-        # Only process every other frame of video to save time
+        # frames = [source.read() for sources in video_sources.values() for source in sources]
-        # Resize frame of video to a smaller size for faster recognition processing
+        for list_of_sources in video_sources.values():
-        run_frame = cv2.resize(frame, (0, 0), fx=args.run_scale, fy=args.run_scale)
+            frames.extend([source.read()[1] for source in list_of_sources])
-        # view_frame = cv2.resize(frame, (0, 0), fx=args.view_scale, fy=args.view_scale)
+        frames_to_show = []
-
+        for frame in frames:
-        results = model(run_frame, verbose=False)
+            frames_to_show.append(
-
+                utils.process_footage(
-        path_to_faces = Path(args.faces_directory)
+                    frame=frame,
-        path_to_faces_exists = path_to_faces.is_dir()
-
-        for i, r in enumerate(results):
-            # list of dicts with each dict containing a label, x1, y1, x2, y2
-            plot_boxes = []
-
-            # The following is stuff for people
-            # This is still in the for loop as each result, no matter if anything is detected, will be present.
-            # Thus, there will always be one result (r)
-
-            # Only run if path_to_faces exists
-            # May be better to check every iteration, but this also works
-            if path_to_faces_exists:
-                if face_details := utils.recognize_face(
-                    path_to_directory=path_to_faces, run_frame=run_frame
-                ):
-                    plot_boxes.append(face_details)
-                    objects_and_peoples = notify.thing_detected(
-                        thing_name=face_details["label"],
-                        objects_and_peoples=objects_and_peoples,
-                        detection_type="peoples",
-                        detection_window=args.detection_window,
-                        detection_duration=args.detection_duration,
-                        notification_window=args.notification_window,
-                        ntfy_url=args.ntfy_url,
-                    )
-
-            # The following is stuff for objects
-            # Setup dictionary of object names
-            if (
-                objects_and_peoples["objects"] == {}
-                or objects_and_peoples["objects"] is None
-            ):
-                for name in r.names.values():
-                    objects_and_peoples["objects"][name] = {
-                        "last_detection_time": None,
-                        "detection_duration": None,
-                        # "first_detection_time": None,
-                        "last_notification_time": None,
-                    }
-                # Also, make sure that the objects to detect are in the list of objects_and_peoples
-                # If it isn't, print a warning
-                for obj in args.detect_object:
-                    if obj not in objects_and_peoples:
-                        print(
-                            f"Warning: {obj} is not in the list of objects the model can detect!"
-                        )
-
-            for box in r.boxes:
-                # Get the name of the object
-                class_id = r.names[box.cls[0].item()]
-                # Get the coordinates of the object
-                cords = box.xyxy[0].tolist()
-                cords = [round(x) for x in cords]
-                # Get the confidence
-                conf = round(box.conf[0].item(), 2)
-                # Print it out, adding a spacer between each object
-                # print("Object type:", class_id)
-                # print("Coordinates:", cords)
-                # print("Probability:", conf)
-                # print("---")
-
-                # Now do stuff (if conf > 0.5)
-                if conf < args.confidence_threshold or (
-                    class_id not in args.detect_object and args.detect_object != []
-                ):
-                    # If the confidence is too low
-                    # or if the object is not in the list of objects to detect and the list of objects to detect is not empty
-                    # then skip this iteration
-                    continue
-
-                # Add the object to the list of objects to plot
-                plot_boxes.append(
-                    {
-                        "label": class_id,
-                        "x1": cords[0],
-                        "y1": cords[1],
-                        "x2": cords[2],
-                        "y2": cords[3],
-                    }
-                )
-
-                objects_and_peoples = notify.thing_detected(
-                    thing_name=class_id,
-                    objects_and_peoples=objects_and_peoples,
-                    detection_type="objects",
-                    detection_window=args.detection_window,
-                    detection_duration=args.detection_duration,
-                    notification_window=args.notification_window,
-                    ntfy_url=args.ntfy_url,
-                )
-
-            # To debug plotting, use r.plot() to cross reference the bounding boxes drawn by the plot_label() and r.plot()
-            frame_to_show = utils.plot_label(
-                boxes=plot_boxes,
-                full_frame=frame,
-                # full_frame=r.plot(),
                     run_scale=args.run_scale,
                     view_scale=args.view_scale,
+                    faces_directory=Path(args.faces_directory),
+                    face_confidence_threshold=args.face_confidence_threshold,
+                    no_remove_representations=args.no_remove_representations,
+                    detection_window=args.detection_window,
+                    detection_duration=args.detection_duration,
+                    notification_window=args.notification_window,
+                    ntfy_url=args.ntfy_url,
+                    model=model,
+                    detect_object=args.detect_object,
+                    object_confidence_threshold=args.object_confidence_threshold,
+                )
             )
-
         # Display the resulting frame
-            # cv2.imshow("", r)
         if not args.no_display:
-                cv2.imshow(f"Video{i}", frame_to_show)
+            for i, frame_to_show in enumerate(frames_to_show):
+                cv2.imshow(f"Video {i}", frame_to_show)
 
         # Hit 'q' on the keyboard to quit!
         if cv2.waitKey(1) & 0xFF == ord("q"):
@@ -314,7 +136,7 @@ def main():
 
     # Release handle to the webcam
     print("Releasing video capture")
-    video_capture.release()
+    [source.release() for sources in video_sources.values() for source in sources]
     cv2.destroyAllWindows()
 
 

wyzely_detect/utils/cli_args.py

@@ -0,0 +1,198 @@
+import argparse
+import os
+import dotenv
+from pathlib import Path
+
+argparser = None
+
+
+def set_argparse():
+    global argparser
+
+    if Path(".env").is_file():
+        dotenv.load_dotenv()
+        print("Loaded .env file")
+    else:
+        print("No .env file found")
+
+    # One important thing to consider is that most function parameters are optional and have a default value
+    # However, with argparse, those are never used since a argparse always passes something, even if it's None
+    argparser = argparse.ArgumentParser(
+        prog="Wyzely Detect",
+        description="Recognize faces/objects in a video stream (from a webcam or a security camera) and send notifications to your devices",  # noqa: E501
+        epilog="For env bool options, setting them to anything except for an empty string will enable them.",
+    )
+
+    video_options = argparser.add_argument_group("Video Options")
+    stream_source = video_options.add_mutually_exclusive_group()
+    stream_source.add_argument(
+        "--rtsp-url",
+        action="append",
+        # If RTSP_URL is in the environment, use it, otherwise just use a blank list
+        # This may cause problems down the road, but if it does, env for this can be removed
+        default=[os.environ["RTSP_URL"]]
+        if "RTSP_URL" in os.environ and os.environ["RTSP_URL"] != ""
+        else [],
+        type=str,
+        help="RTSP camera URL",
+    )
+    stream_source.add_argument(
+        "--capture-device",
+        action="append",
+        # If CAPTURE_DEVICE is in the environment, use it, otherwise just use a blank list
+        # If __main__.py detects that no capture device or remote stream is set, it will default to 0
+        default=[int(os.environ["CAPTURE_DEVICE"])]
+        if "CAPTURE_DEVICE" in os.environ and os.environ["CAPTURE_DEVICE"] != ""
+        else [],
+        type=int,
+        help="Capture device number",
+    )
+    video_options.add_argument(
+        "--run-scale",
+        # Set it to the env RUN_SCALE if it isn't blank, otherwise set it to 0.25
+        default=os.environ["RUN_SCALE"]
+        if "RUN_SCALE" in os.environ and os.environ["RUN_SCALE"] != ""
+        # else 0.25,
+        else 1,
+        type=float,
+        help="The scale to run the detection at, default is 0.25",
+    )
+    video_options.add_argument(
+        "--view-scale",
+        # Set it to the env VIEW_SCALE if it isn't blank, otherwise set it to 0.75
+        default=os.environ["VIEW_SCALE"]
+        if "VIEW_SCALE" in os.environ and os.environ["VIEW_SCALE"] != ""
+        # else 0.75,
+        else 1,
+        type=float,
+        help="The scale to view the detection at, default is 0.75",
+    )
+
+    video_options.add_argument(
+        "--no-display",
+        default=os.environ["NO_DISPLAY"]
+        if "NO_DISPLAY" in os.environ
+        and os.environ["NO_DISPLAY"] != ""
+        and os.environ["NO_DISPLAY"].lower() != "false"
+        else False,
+        action="store_true",
+        help="Don't display the video feed",
+    )
+    video_options.add_argument(
+        "-c",
+        "--force-disable-tensorflow-gpu",
+        default=os.environ["FORCE_DISABLE_TENSORFLOW_GPU"]
+        if "FORCE_DISABLE_TENSORFLOW_GPU" in os.environ
+        and os.environ["FORCE_DISABLE_TENSORFLOW_GPU"] != ""
+        and os.environ["FORCE_DISABLE_TENSORFLOW_GPU"].lower() != "false"
+        else False,
+        action="store_true",
+        help="Force disable tensorflow GPU through env since sometimes it's not worth it to install cudnn and whatnot",
+    )
+
+    notifcation_services = argparser.add_argument_group("Notification Services")
+    notifcation_services.add_argument(
+        "--ntfy-url",
+        default=os.environ["NTFY_URL"]
+        if "NTFY_URL" in os.environ and os.environ["NTFY_URL"] != ""
+        # This is None but there is a default set in notify.py
+        else None,
+        type=str,
+        help="The URL to send notifications to",
+    )
+
+    # Various timers
+    timers = argparser.add_argument_group("Timers")
+    timers.add_argument(
+        "--detection-duration",
+        default=os.environ["DETECTION_DURATION"]
+        if "DETECTION_DURATION" in os.environ and os.environ["DETECTION_DURATION"] != ""
+        else 2,
+        type=int,
+        help="The duration (in seconds) that an object must be detected for before sending a notification",
+    )
+    timers.add_argument(
+        "--detection-window",
+        default=os.environ["DETECTION_WINDOW"]
+        if "DETECTION_WINDOW" in os.environ and os.environ["DETECTION_WINDOW"] != ""
+        else 15,
+        type=int,
+        help="The time (seconds) before the detection duration resets",
+    )
+    timers.add_argument(
+        "--notification-window",
+        default=os.environ["NOTIFICATION_WINDOW"]
+        if "NOTIFICATION_WINDOW" in os.environ
+        and os.environ["NOTIFICATION_WINDOW"] != ""
+        else 30,
+        type=int,
+        help="The time (seconds) before another notification can be sent",
+    )
+
+    face_recognition = argparser.add_argument_group("Face Recognition options")
+    face_recognition.add_argument(
+        "--faces-directory",
+        default=os.environ["FACES_DIRECTORY"]
+        if "FACES_DIRECTORY" in os.environ and os.environ["FACES_DIRECTORY"] != ""
+        else "faces",
+        type=str,
+        help="The directory to store the faces. Can either contain images or subdirectories with images, the latter being the preferred method",  # noqa: E501
+    )
+    face_recognition.add_argument(
+        "--face-confidence-threshold",
+        default=os.environ["FACE_CONFIDENCE_THRESHOLD"]
+        if "FACE_CONFIDENCE_THRESHOLD" in os.environ
+        and os.environ["FACE_CONFIDENCE_THRESHOLD"] != ""
+        else 0.3,
+        type=float,
+        help="The confidence (currently cosine similarity) threshold to use for face recognition",
+    )
+    face_recognition.add_argument(
+        "--no-remove-representations",
+        default=os.environ["NO_REMOVE_REPRESENTATIONS"]
+        if "NO_REMOVE_REPRESENTATIONS" in os.environ
+        and os.environ["NO_REMOVE_REPRESENTATIONS"] != ""
+        and os.environ["NO_REMOVE_REPRESENTATIONS"].lower() != "false"
+        else False,
+        action="store_true",
+        help="Don't remove representations_<model>.pkl at the start of the program. Greatly improves startup time, but doesn't take into account changes to the faces directory since it was created",  # noqa: E501
+    )
+
+    object_detection = argparser.add_argument_group("Object Detection options")
+    object_detection.add_argument(
+        "--detect-object",
+        action="append",
+        # Stuff is appended to default, as far as I can tell
+        default=[],
+        type=str,
+        help="The object(s) to detect. Must be something the model is trained to detect",
+    )
+    object_detection.add_argument(
+        "--object-confidence-threshold",
+        default=os.environ["OBJECT_CONFIDENCE_THRESHOLD"]
+        if "OBJECT_CONFIDENCE_THRESHOLD" in os.environ
+        and os.environ["OBJECT_CONFIDENCE_THRESHOLD"] != ""
+        # I think this should always be a str so using lower shouldn't be a problem.
+        # Also, if the first check fails the rest shouldn't be run
+        and os.environ["OBJECT_CONFIDENCE_THRESHOLD"].lower() != "false" else 0.6,
+        type=float,
+        help="The confidence threshold to use",
+    )
+
+    debug = argparser.add_argument_group("Debug options")
+    debug.add_argument(
+        "--fake-second-source",
+        help="Duplicate the first source and use it as a second source. Capture device takes priority.",
+        action="store_true",
+        default=os.environ["FAKE_SECOND_SOURCE"]
+        if "FAKE_SECOND_SOURCE" in os.environ
+        and os.environ["FAKE_SECOND_SOURCE"] != ""
+        and os.environ["FAKE_SECOND_SOURCE"].lower() != "false"
+        else False,
+    )
+
+    # return argparser
+
+
+# This will run when this file is imported
+set_argparse()

wyzely_detect/utils/notify.py

@@ -104,6 +104,11 @@ def thing_detected(
         ):
             respective_type[thing_name]["last_notification_time"] = time.time()
             print(f"Detected {thing_name} for {detection_duration} seconds")
+            if ntfy_url is None:
+                print(
+                    "ntfy_url is None. Not sending notification. Set ntfy_url to send notifications"
+                )
+            else:
                 headers = construct_ntfy_headers(
                     title=f"{thing_name} detected",
                     tag="rotating_light",

wyzely_detect/utils/utils.py

@@ -1,10 +1,163 @@
 import cv2
+import os
 import numpy as np
 from pathlib import Path
-from deepface import DeepFace
+
+# https://stackoverflow.com/a/42121886/18270659
+os.environ["TF_CPP_MIN_LOG_LEVEL"] = "3"
+
+
+from deepface import DeepFace  # noqa: E402
+from . import notify  # noqa: E402
 
 first_face_try = True
 
+# TODO: When multi-camera support is ~~added~~ improved, this will need to be changed so that each camera has its own dict
+objects_and_peoples = {
+    "objects": {},
+    "peoples": {},
+}
+
+
+def process_footage(
+    # Frame
+    frame: np.ndarray = None,
+    # scale
+    run_scale: float = None,
+    view_scale: float = None,
+    # Face stuff
+    faces_directory: str = None,
+    face_confidence_threshold: float = None,
+    no_remove_representations: bool = False,
+    # Timer stuff
+    detection_window: int = None,
+    detection_duration: int = None,
+    notification_window: int = None,
+    ntfy_url: str = None,
+    # Object stuff
+    # YOLO object
+    model=None,
+    detect_object: list = None,
+    object_confidence_threshold=None,
+) -> np.ndarray:
+    """Takes in a frame and processes it"""
+    global objects_and_peoples
+
+    # Resize frame of video to a smaller size for faster recognition processing
+    run_frame = cv2.resize(frame, (0, 0), fx=run_scale, fy=run_scale)
+    # view_frame = cv2.resize(frame, (0, 0), fx=args.view_scale, fy=args.view_scale)
+
+    results = model(run_frame, verbose=False)
+
+    path_to_faces = Path(faces_directory)
+    path_to_faces_exists = path_to_faces.is_dir()
+
+    for r in results:
+        # list of dicts with each dict containing a label, x1, y1, x2, y2
+        plot_boxes = []
+
+        # The following is stuff for people
+        # This is still in the for loop as each result, no matter if anything is detected, will be present.
+        # Thus, there will always be one result (r)
+
+        # Only run if path_to_faces exists
+        # May be better to check every iteration, but this also works
+        if path_to_faces_exists:
+            if face_details := recognize_face(
+                path_to_directory=path_to_faces,
+                run_frame=run_frame,
+                # Perhaps make these names match?
+                min_confidence=face_confidence_threshold,
+                no_remove_representations=no_remove_representations,
+            ):
+                plot_boxes.append(face_details)
+                objects_and_peoples = notify.thing_detected(
+                    thing_name=face_details["label"],
+                    objects_and_peoples=objects_and_peoples,
+                    detection_type="peoples",
+                    detection_window=detection_window,
+                    detection_duration=detection_duration,
+                    notification_window=notification_window,
+                    ntfy_url=ntfy_url,
+                )
+
+        # The following is stuff for objects
+        # Setup dictionary of object names
+        if (
+            objects_and_peoples["objects"] == {}
+            or objects_and_peoples["objects"] is None
+        ):
+            for name in r.names.values():
+                objects_and_peoples["objects"][name] = {
+                    "last_detection_time": None,
+                    "detection_duration": None,
+                    # "first_detection_time": None,
+                    "last_notification_time": None,
+                }
+            # Also, make sure that the objects to detect are in the list of objects_and_peoples
+            # If it isn't, print a warning
+            for obj in detect_object:
+                # .keys() shouldn't be needed
+                if obj not in objects_and_peoples["objects"]:
+                    print(
+                        f"Warning: {obj} is not in the list of objects the model can detect!"
+                    )
+
+        for box in r.boxes:
+            # Get the name of the object
+            class_id = r.names[box.cls[0].item()]
+            # Get the coordinates of the object
+            cords = box.xyxy[0].tolist()
+            cords = [round(x) for x in cords]
+            # Get the confidence
+            conf = round(box.conf[0].item(), 2)
+            # Print it out, adding a spacer between each object
+            # print("Object type:", class_id)
+            # print("Coordinates:", cords)
+            # print("Probability:", conf)
+            # print("---")
+
+            # Now do stuff (if conf > 0.5)
+            if conf < object_confidence_threshold or (
+                class_id not in detect_object and detect_object != []
+            ):
+                # If the confidence is too low
+                # or if the object is not in the list of objects to detect and the list of objects to detect is not empty
+                # then skip this iteration
+                continue
+
+            # Add the object to the list of objects to plot
+            plot_boxes.append(
+                {
+                    "label": class_id,
+                    "x1": cords[0],
+                    "y1": cords[1],
+                    "x2": cords[2],
+                    "y2": cords[3],
+                }
+            )
+
+            objects_and_peoples = notify.thing_detected(
+                thing_name=class_id,
+                objects_and_peoples=objects_and_peoples,
+                detection_type="objects",
+                detection_window=detection_window,
+                detection_duration=detection_duration,
+                notification_window=notification_window,
+                ntfy_url=ntfy_url,
+            )
+
+        # To debug plotting, use r.plot() to cross reference the bounding boxes drawn by the plot_label() and r.plot()
+        frame_to_show = plot_label(
+            boxes=plot_boxes,
+            full_frame=frame,
+            # full_frame=r.plot(),
+            run_scale=run_scale,
+            view_scale=view_scale,
+        )
+        # Unsure if this should also return the objects_and_peoples dict
+        return frame_to_show
+
 
 def plot_label(
     # list of dicts with each dict containing a label, x1, y1, x2, y2
@@ -18,7 +171,7 @@ def plot_label(
     # So the coordinates will be scaled appropriately when coming from run_frame
     view_scale: float = None,
     font: int = cv2.FONT_HERSHEY_SIMPLEX,
-):
+) -> np.ndarray:
     # x1 and y1 are the top left corner of the box
     # x2 and y2 are the bottom right corner of the box
     # Example scaling: full_frame: 1 run_frame: 0.5 view_frame: 0.25
@@ -68,6 +221,8 @@ def recognize_face(
     path_to_directory: Path = Path("faces"),
     # opencv image
     run_frame: np.ndarray = None,
+    min_confidence: float = 0.3,
+    no_remove_representations: bool = False,
 ) -> np.ndarray:
     """
         Accepts a path to a directory of images of faces to be used as a refference
@@ -75,7 +230,8 @@ def recognize_face(
 
         Returns a single dictonary as currently only 1 face can be detected in each frame
         Cosine threshold is 0.3, so if the confidence is less than that, it will return None
-    dict contains the following keys: label, x1, y1, x2, y2
+        dict conta                # Maybe use os.exit() instead?
+    ins the following keys: label, x1, y1, x2, y2
         The directory should be structured as follows:
         faces/
             name/
@@ -94,13 +250,16 @@ def recognize_face(
     global first_face_try
 
     # If it's the first time the function is being run, remove representations_arcface.pkl, if it exists
-    if first_face_try:
+    if first_face_try and not no_remove_representations:
         try:
             path_to_directory.joinpath("representations_arcface.pkl").unlink()
             print("Removing representations_arcface.pkl")
         except FileNotFoundError:
             print("representations_arcface.pkl does not exist")
         first_face_try = False
+    elif first_face_try and no_remove_representations:
+        print("Not attempting to remove representations_arcface.pkl")
+        first_face_try = False
 
     # face_dataframes is a vanilla list of dataframes
     # It seems face_dataframes is empty if the face database (directory) doesn't exist. Seems to work if it's empty though
@@ -119,6 +278,10 @@ def recognize_face(
             model_name="ArcFace",
             detector_backend="opencv",
         )
+        '''
+        Example dataframe, for reference
+        identity  (path to image) | source_x | source_y | source_w | source_h | VGG-Face_cosine (pretty much the confidence \\_('_')_/) 
+        '''
     except ValueError as e:
         if (
             str(e)
@@ -126,6 +289,14 @@ def recognize_face(
         ):
             # print("No faces recognized") # For debugging
             return None
+        elif (
+            # Check if the error message contains "Validate .jpg or .png files exist in this path."
+            "Validate .jpg or .png files exist in this path."
+            in str(e)
+        ):
+            # If a verbose/silent flag is added, this should be changed to print only if verbose is true
+            # print("No faces found in database")
+            return None
         else:
             raise e
     # Iteate over the dataframes
@@ -133,8 +304,13 @@ def recognize_face(
         # The last row is the highest confidence
         # So we can just grab the path from there
         # iloc = Integer LOCation
+        try:
             path_to_image = Path(df.iloc[-1]["identity"])
-        # If the parent name is the same as the path to the database, then set label to the image name instead of the parent directory name
+        # Seems this is caused when someone steps into frame and their face is detected but not recognized
+        except IndexError:
+            print("Face present but not recognized")
+            continue
+        # If the parent name is the same as the path to the database, then set label to the image name instead of the parent name
         if path_to_image.parent == Path(path_to_directory):
             label = path_to_image.name
         else:
@@ -149,19 +325,13 @@ def recognize_face(
             "y2": df.iloc[-1]["source_y"] + df.iloc[-1]["source_h"],
         }
         # After some brief testing, it seems positive matches are > 0.3
-        distance = df.iloc[-1]["ArcFace_cosine"]
+        cosine_similarity = df.iloc[-1]["ArcFace_cosine"]
-        # TODO: Make this a CLI argument
+        if cosine_similarity < min_confidence:
-        if distance < 0.3:
             return None
-        # if 0.5 < distance < 0.7:
         # label = "Unknown"
         to_return = dict(label=label, **coordinates)
         print(
-            f"Confindence: {distance}, filname: {path_to_image.name}, to_return: {to_return}"
+            f"Cosine similarity: {cosine_similarity}, filname: {path_to_image.name}, to_return: {to_return}"
         )
         return to_return
-
+    return None
-    """
-    Example dataframe, for reference
-    identity  (path to image) | source_x | source_y | source_w | source_h | VGG-Face_cosine (pretty much the confidence \_('_')_/) 
-    """