Image attribute recognition code implementation based on ALM model [1]
pip install -r requirements.txt
- Download the trained model from here.
- Put the model
saved_model.pt
file under./model
or./app/alm/model
- Train
python main.py --experiment=rap
- Visualize
Start TensorBoard with
python main.py --experiment=rap --vis
tensorboard --logdir ./runs
[1] Tang, C., Sheng, L., Zhang, Z., & Hu, X. (2019). Improving pedestrian attribute recognition with weakly-supervised multi-scale attribute-specific localization. In Proceedings of the IEEE/CVF International Conference on Computer Vision (pp. 4997-5006).