Sign Language Project

Deploying a YOLOv7 detecting American Sign Language (ASL) on Intel RealSense

Scripts:

Training YOLOv7:

Open the notebook:

Sign_Language_Training_on_Custom_Data.ipynb

!python train.py --batch 8 --epochs 200 --data /yolov7/American-Sign-Language-Letters/data.yaml \
--weights '/yolov7/runs/train/exp11/weights/best.pt' \
--device 0 --img 640 640 --cfg '/cfg/training/yolov7.yaml'

Training Modified ResNet18:

image_classification.ipynb

Testing yolov7 - Camera or Image upload:

test_yolo.py

Testing Modified Resnet - Camera or Image upload:

test_resnet.py

Required Libraries:

(Note some of them might not be needed depending on the application)

inside yolov7 jupyter notebook, run the command to install dependencies from Requirements.txt.

pyrealsense2

pip install pyrealsense2

Install requests

pip3 install requests

Install Pillows for PIL image

pip install Pillow==6.1

Pandas

pip install pandas

Pytorch

pip3 install torch torchvision torchaudio

tqdm

pip install tqdm

pyyaml

pip install pyyaml

Misc

python -m pip install -U matplotlib
pip install seaborn
pip install scipy

For exporting Model:

pip install nvidia-pyindex
pip install --upgrade nvidia-tensorrt
pip install pycuda

pip install protobuf<4.21.3
pip install onnxruntime-gpu
pip install onnx>=1.9.0
pip install onnx-simplifier>=0.3.6

pip install nvidia-pyindex
pip install onnx-graphsurgeon

To use Jupyter Notebooks

pip install jupyter
pip install ipywidgets widgetsnbextension pandas-profiling