Abstract:Dense object detection is widely used in automatic driving, video surveillance, and other fields. This paper focuses on the challenging task of dense object detection. Currently, detection methods based on greedy algorithms, such as non-maximum suppression (NMS), often produce many repetitive predictions or missed detections in dense scenarios, which is a common problem faced by NMS-based algorithms. Through the end-to-end DETR (DEtection TRansformer), as a type of detector that can incorporate the post-processing de-duplication capability of NMS, etc., into the network, we found that homogeneous queries in the query-based detector lead to a reduction in the de-duplication capability of the network and the learning efficiency of the encoder, resulting in duplicate prediction and missed detection problems. To solve this problem, we propose learnable differentiated encoding to de-homogenize the queries, and at the same time, queries can communicate with each other via differentiated encoding information, replacing the previous self-attention among the queries. In addition, we used joint loss on the output of the encoder that considered both location and confidence prediction to give a higher-quality initialization for queries. Without cumbersome decoder stacking and guaranteeing accuracy, our proposed end-to-end detection framework was more concise and reduced the number of parameters by about 8% compared to deformable DETR. Our method achieved excellent results on the challenging CrowdHuman dataset with 93.6% average precision (AP), 39.2% MR−2, and 84.3% JI. The performance overperformed previous SOTA methods, such as Iter-E2EDet (Progressive End-to-End Object Detection) and MIP (One proposal, Multiple predictions). In addition, our method is more robust in various scenarios with different densities.

Can the Query-based Object Detector Be Designed with Fewer Stages?

StageInteractor: Query-based Object Detector with Cross-stage Interaction

A Lightweight SE-YOLOv3 Network for Multi-Scale Object Detection in Remote Sensing Imagery.

Enhanced Training of Query-Based Object Detection via Selective Query Recollection

An Object Detection Method Based on Improved YOLOX

DECO: Query-Based End-to-End Object Detection with ConvNets

Dq-YOLOF: An Effective Improvement with Deformable Convolution and Sample Quality Optimization Based on the YOLOF Detector

Efficient DETR: Improving End-to-End Object Detector with Dense Prior

Redundant Queries in DETR-Based 3D Detection Methods: Unnecessary and Prunable

Towards Discriminative and Transferable One-Stage Few-Shot Object Detectors

Efficient One-stage Video Object Detection by Exploiting Temporal Consistency

Dense Object Detection Based on De-Homogenized Queries

Deep Equilibrium Object Detection

Guided Refine-Head For Object Detection

Featurized Query R-CNN

Anchor DETR: Query Design for Transformer-Based Detector

QueryDet: Cascaded Sparse Query for Accelerating High-Resolution Small Object Detection

Multi-modal Queried Object Detection in the Wild

Condensing Two-stage Detection with Automatic Object Key Part Discovery

Dynamic Object Queries for Transformer-based Incremental Object Detection

What Are Expected Queries in End-to-End Object Detection?