title	emoji	sdk	app_port	pinned
lol_highlight_detection	🐳	docker	8765	false

lol_highlight_detection

League of Legend Highlight Detection

Running project

1. Use devcontainer, should be automatic
2. Use conda/mamba and install env.yml, simple enough!
3. solara run sol_app.py

Presentation

slides.google.com

Resources:

1. Fast.AI - https://docs.fast.ai/tutorial.image_sequence.html
2. TIMM/HF - https://github.com/huggingface/pytorch-image-models
3. HF/ViT - see video_classification.ipynb
4. VideoMAE - https://huggingface.co/docs/transformers/model_doc/videomae
5. Keras - https://keras.io/examples/vision/video_classification/ (keras-core??)
6. TF - https://www.tensorflow.org/tutorials/video/video_classification
7. Papers - https://paperswithcode.com/search?q_meta=&q_type=&q=videomae
8. Hiera - https://github.com/facebookresearch/hiera