Releases: vita-epfl/bounding-box-prediction
Releases · vita-epfl/bounding-box-prediction
Multi-Task PV-LSTM Trained
Multi-Task PV-LSTM trained on JAAD dataset using the following setting:
- Input sequence length = 16 frames (0.53 sec)
- Output sequence length = 16 frames (0.53 sec)
- Batch size = 100
- number of epochs = 100