Implemented the Transformer Technology on selective classes for Human Gait Recognition
- I have implemented the Vision Transformer to train on the sequence of the Human Poses
- These Human poses are extracted from the HRNet.
- Initially, this vision transformer is used for image recognition by passing the sequence of patches of images
- I changed the model to learn the sequence of Human Poses that will result in Human Gait Recognition
- Please use the embedding layer exactly equal to the sequence length
- Please make sure to
Query
andValues
. As these are both required for the Transformer
Feel free to contact me @ abilal.bee20seecs@seecs.edu.pk