Bo Dai

Hi, The input consists of two h5 files. One file ('data') contains images, sequences and image indexes of the sequences. Another file ('feat') contains features of the images. h5ls 'data':...

@cbsudux @Happymarrow sorry for the late reply. coco_cap_mappings.json contains to mappings {"wtoi": {}, "itow": {}}, which are used for word to index and index to word respectively. You can refer...

This repo does not include code for object detection since no modification is made to existing code. You may refer to repo of Faster-RCNN, and rearrange the detection results into...

I further cleaned the dataset, I think it's better to refer to this released one in your projects.

I think I've used the version 1.0. I removed redundant bounding boxes but didn't move bounding boxes.