happywu / Sequence-Level-Semantics-Aggregation

Sequence Level Semantics Aggregation for Video Object Detection
Apache License 2.0
85 stars 19 forks source link

Could you provide detailed introductions or explanation for the meanings of txt file inside /ImageSets/ folder?Thanks #4

Closed TyroneLi closed 4 years ago

happywu commented 4 years ago

For video data (VID), there is four fields: name /path of the video segment, start_frame_id, frame_idx, total length of the segment For single frame data (DET), there is two fields: path of the image, start_frame_id (for consistent / convenience with video data) You shall refer to https://github.com/happywu/Sequence-Level-Semantics-Aggregation/blob/7ae05968bb9a7e3da3ca9722d1d13fe6306ffdbd/lib/dataset/imagenet_vid.py#L70-L87 for usage.