Video Camera deep learning to Detect Animal Behavior
Budget: $750 – $1,500 USD
1. Data processing: You will connect to bunch of video cameras and need to run ML/Deep learning to analyze the video frames for specific behaviors of animals and nearby, for example: Animal laying, standing, eating, walking, lame, person passing, vehicle pass by etc.
2. Video Annotation: need easy framework to custom label / annotate the video fast and be able to tag fixed events but also behavior
3. All data should be then counted/aggregated and stored with timestamp and behavior in database and reporting tool (e.g. google sheets)
Please note:
1. several cameras and each can be in different tilt, so custom annotation and training should be allowed for that
2. Need to choose best and low-code drag n drop framework but that is cheap (e.g. YOLO, tensor flow other) + cheap hosting
3. all need to be simple drag n drop low code, and cheap to host and process (which cloud platform?)
The project includes:
1. Assessing with me the best platform and framework and tool to use and price prediction for 24x7 video analysis
2. Help in labeling/annotating the videos to the behavior and tagging I pre-define
3. train the mode
4. Test model on several cameras for accuracy
5. Store results in straight or aggregated manner in database and mirror to google sheet / reporting tool
6. Package the system to a run time host of cheap cloud and let it run 24x7
7. optimize
2. Video Annotation: need easy framework to custom label / annotate the video fast and be able to tag fixed events but also behavior
3. All data should be then counted/aggregated and stored with timestamp and behavior in database and reporting tool (e.g. google sheets)
Please note:
1. several cameras and each can be in different tilt, so custom annotation and training should be allowed for that
2. Need to choose best and low-code drag n drop framework but that is cheap (e.g. YOLO, tensor flow other) + cheap hosting
3. all need to be simple drag n drop low code, and cheap to host and process (which cloud platform?)
The project includes:
1. Assessing with me the best platform and framework and tool to use and price prediction for 24x7 video analysis
2. Help in labeling/annotating the videos to the behavior and tagging I pre-define
3. train the mode
4. Test model on several cameras for accuracy
5. Store results in straight or aggregated manner in database and mirror to google sheet / reporting tool
6. Package the system to a run time host of cheap cloud and let it run 24x7
7. optimize