Cross-DINO: Cross the Deep MLP and Transformer for Small Object Detection

TL;DR

Cross-DINO introduces a novel method combining deep MLP and Transformer encoders with a new module and loss function to significantly improve small object detection performance in various datasets.

Contribution

The paper proposes Cross-DINO, a new approach that integrates deep MLP, a Cross Coding Twice Module, and a Boost Loss with a new soft label for enhanced small object detection.

Findings

Achieves 36.4% APs on COCO for SOD, outperforming DINO by 4.4%.

Requires only 45M parameters, fewer FLOPs, and under 12 epochs training.

Demonstrates effectiveness across multiple datasets like COCO, WiderPerson, VisDrone, AI-TOD, and SODA-D.

Abstract

Small Object Detection (SOD) poses significant challenges due to limited information and the model's low class prediction score. While Transformer-based detectors have shown promising performance, their potential for SOD remains largely unexplored. In typical DETR-like frameworks, the CNN backbone network, specialized in aggregating local information, struggles to capture the necessary contextual information for SOD. The multiple attention layers in the Transformer Encoder face difficulties in effectively attending to small objects and can also lead to blurring of features. Furthermore, the model's lower class prediction score of small objects compared to large objects further increases the difficulty of SOD. To address these challenges, we introduce a novel approach called Cross-DINO. This approach incorporates the deep MLP network to aggregate initial feature representations with both short and long range information for SOD. Then, a new Cross Coding Twice Module (CCTM) is applied to integrate these initial representations to the Transformer Encoder feature, enhancing the details of small objects. Additionally, we introduce a new kind of soft label named Category-Size (CS), integrating the Category and Size of objects. By treating CS as new ground truth, we propose a new loss function called Boost Loss to improve the class prediction score of the model. Extensive experimental results on COCO, WiderPerson, VisDrone, AI-TOD, and SODA-D datasets demonstrate that Cross-DINO efficiently improves the performance of DETR-like models on SOD. Specifically, our model achieves 36.4% APs on COCO for SOD with only 45M parameters, outperforming the DINO by +4.4% APS (36.4% vs. 32.0%) with fewer parameters and FLOPs, under 12 epochs training setting. The source codes will be available at https://github.com/Med-Process/Cross-DINO.