Mr. DETR: Instructive Multi-Route Training for Detection Transformers
Paper: Mr. DETR: Instructive Multi-Route Training for Detection Transformers
Project Page: https://visual-ai.github.io/mrdetr
Code: https://github.com/Visual-AI/Mr.DETR
Abstract
Existing methods enhance the training of detection transformers by incorporating an auxiliary one-to-many assignment. In this work, we treat the model as a multi-task framework, simultaneously performing one-to-one and one-to-many predictions. We investigate the roles of each component in the transformer decoder across these two training targets, including self-attention, cross-attention, and feed-forward network. Our empirical results demonstrate that any independent component in the decoder can effectively learn both targets simultaneously, even when other components are shared. This finding leads us to propose a multi-route training mechanism, featuring a primary route for one-to-one prediction and two auxiliary training routes for one-to-many prediction. We enhance the training mechanism with a novel instructive self-attention that dynamically and flexibly guides object queries for one-to-many prediction. The auxiliary routes are removed during inference, ensuring no impact on model architecture or inference cost. We conduct extensive experiments on various baselines, achieving consistent improvements as shown in Figure 1.
Updates
- [04/25] Mr. DETR supports Instance segmentation now. We release the code and pre-trained weights.
- [03/25] We release the code and weights of Mr. DETR for object detection. You may find pre-trained weights at Huggingface.
- [03/25] Mr. DETR is accepted by CVPR 2025.