Any way to use the original images

MILVLG / openvqa

A lightweight, scalable, and general framework for visual question answering research

Apache License 2.0

320 stars 64 forks source link

Any way to use the original images #76

Closed Dhruv-Sabharwal closed 2 years ago

Dhruv-Sabharwal commented 2 years ago

Hi, I was implementing some models to do VQA and found your repo really useful. However, it seems like I can only get the FRCN_FEAT and BBOX_FEAT as image inputs to the model. Is there any way to take the original images as inputs and not the extracted features?

MIL-VLG commented 2 years ago

It is not that easy and needs plenty of modifications. Currently we do not have such plan to support this.