A model which detects and interprets sign language in real time which reduces the communications barrier which exists between the deaf and hearing communities.
- Python using Anaconda
- Visual C++ Build Tools 2022 version
- CUDA and CUDNN for Nvidia
- Protoc
- Tensorflow Object Detection API
Note : Make sure to check desktop development with C++ while installing
-
Install tensorflow model Tensorflow Models GitHub
-
Open
cmd
--> gotomodels\research
folder andenter the following command
protoc object_detection/protos/*.proto --python_out =.
-
Copy the
setup.py folder
fromobject_detection/packages/tf2 folder
into theresearch folder
-
To install all the dependencies needed for the object detection library, run
python -m pip install .
We trained the model for 10000 steps
Python Tensorflow/models/research/object_detection/model_main_tv2.py --model_dir = Tensorflow/workspace/models/my_ssd_ mobnet --
pipeline_config_path = Tensorflow/workspace/models/my_ssd_mobnet/pipeline.config --num_train_steps = 10000
Jihyeok Choi
Hyelim Yang
Pragati Singh
Rohit Chandra