31 lines (24 loc) · 886 Bytes

title	lol_highlight_detection
emoji	🐳
sdk	docker
app_port	8765
pinned	false

lol_highlight_detection

League of Legend Highlight Detection

Running project

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

Presentation

slides.google.com

Resources:

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