msracver / Flow-Guided-Feature-Aggregation

Flow-Guided Feature Aggregation for Video Object Detection
MIT License
723 stars 190 forks source link

How can I perform mixed training with DET and VID datasets, considering that the DET dataset does not consist of consecutive frames? #97

Open KaihongLi opened 5 months ago

KaihongLi commented 5 months ago

How can I perform mixed training with DET and VID datasets, considering that the DET dataset does not consist of consecutive frames? In your code, you mentioned performing only a single-stage mixed training. How can the DET dataset be incorporated into this mixed training approach?