MILVLG / bottom-up-attention.pytorch

A PyTorch reimplementation of bottom-up-attention models
Apache License 2.0
294 stars 76 forks source link

how to extract features on my own dataset #72

Closed Chunmian-art closed 3 years ago

Chunmian-art commented 3 years ago

Hi, thank you for the nice work! But I still have some questions on extract features on my own dataset. I train an object detection model through detectron2. But how can I use this model in your work? Or should I use your work train an object detection model?

Zoroaster97 commented 3 years ago

This framework is not support training models yet (we are still working on it 😢). We provide some pre-trained models, you can use these models to extract features on your datasets. If you want to use your model, you can try to modify the "model" part of this framework, and replace it with your code.

leonodelee commented 2 years ago

This framework is not support training models yet (we are still working on it 😢). We provide some pre-trained models, you can use these models to extract features on your datasets. If you want to use your model, you can try to modify the "model" part of this framework, and replace it with your code.

I wonder if I can use my own model.pth to extract features on my datasets using your codebase?