Patch-based Selection and Refinement for Early Object Detection

Tianyi Zhang, Kishore Kasichainula, Yaoxin Zhuo, Baoxin Li, Jae-Sun Seo +1 more
11/3/2023
cs.CV

Official code: destiny301/dpr (22 stars, last commit 2024-04-19).

Abstract

Early object detection (OD) is a crucial task for the safety of many dynamic systems. Current OD algorithms have limited success for small objects at a long distance. To improve the accuracy and efficiency of such a task, we propose a novel set of algorithms that divide the image into patches, select patches with objects at various scales, elaborate the details of a small object, and detect it as early as possible. Our approach is built upon a transformer-based network and integrates the diffusion model to improve the detection accuracy. As demonstrated on BDD100K, our algorithms enhance the mAP for small objects from 1.03 to 8.93, and reduce the data volume in computation by more than 77\%. The source code is available at \href{https://github.com/destiny301/dpr}{https://github.com/destiny301/dpr}

View on arXivView PDF

Code Implementations(1)

destiny301/dprOfficial100%

[WACV2024] Patch-based Selection and Refinement for Early Object Detection

223PythonJul 17, 20232 years ago

Similar Papers

A Real-Time DETR Approach to Bangladesh Road Object Detection for Autonomous Vehicles

88% match2024

Transformers in Small Object Detection: A Benchmark and Survey of State-of-the-Art

86% match1 repo1722023

YOLO-Former: YOLO Shakes Hand With ViT

86% match2024

SOD-YOLOv8 -- Enhancing YOLOv8 for Small Object Detection in Traffic Scenes

85% match1 repo452024

Contrastive Learning for Multi-Object Tracking with Transformers

85% match1 repo92023

Small Object Detection by DETR via Information Augmentation and Adaptive Feature Fusion

85% match2024

Cite this paper

@article{zhang2023patchbased,
  title  = {Patch-based Selection and Refinement for Early Object Detection},
  author = {Tianyi Zhang and Kishore Kasichainula and Yaoxin Zhuo and Baoxin Li and Jae-Sun Seo and Yu Cao},
  year   = {2023},
  eprint = {2311.02274},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2311.02274v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2311.02274.svg)](https://codeofpaper.com/paper/2311.02274)

Discussion