Project Description


Hand gesture recognition is very significant for human-computer
interaction. In this work, we present a novel real-time method for hand
gesture recognition. In this project, we propose to design a real time system
to detect the hand gesture for the Computer System. We
propose to use the techniques of machine learning and computer vision to
perform hand gesture recognition. We propose to collect multiple images of
different and distinct classes of hand gestures and propose to design a
transfer learning based deep learning system to classify the images.


Run and Installation



To Perform Gesture Recognition, Execute each of the lines one after the other in terminal

Training: (One time)

IMAGE_SIZE=224

ARCHITECTURE="mobilenet_1.0_${IMAGE_SIZE}"

pkill -f "tensorboard"

tensorboard --logdir tf_files/training_summaries &

python -m scripts.retrain \
  --bottleneck_dir=tf_files/bottlenecks \
  --how_many_training_steps=500 \
  --model_dir=tf_files/models/ \
  --summaries_dir=tf_files/training_summaries/"${ARCHITECTURE}" \
  --output_graph=tf_files/retrained_graph.pb \
  --output_labels=tf_files/retrained_labels.txt \
  --architecture="${ARCHITECTURE}" \
  --image_dir=tf_files/Gesture_images

Testing:

chmod +x run_test.sh
./run_test.sh