A Solution to Product detection in Densely Packed Scenes

Tianze Rong, Yanjia Zhu, Hongxiang Cai, Yichao Xiong

2020-07-23Dense Object Detection

Abstract

This work is a solution to densely packed scenes dataset SKU-110k. Our work is modified from Cascade R-CNN. To solve the problem, we proposed a random crop strategy to ensure both the sampling rate and input scale is relatively sufficient as a contrast to the regular random crop. And we adopted some of trick and optimized the hyper-parameters. To grasp the essential feature of the densely packed scenes, we analysis the stages of a detector and investigate the bottleneck which limits the performance. As a result, our method obtains 58.7 mAP on test set of SKU-110k.

Results

Task	Dataset	Metric	Value	Model
Object Detection	SKU-110K	AP	58.7	Cascade-RCNN
3D	SKU-110K	AP	58.7	Cascade-RCNN
2D Classification	SKU-110K	AP	58.7	Cascade-RCNN
2D Object Detection	SKU-110K	AP	58.7	Cascade-RCNN
16k	SKU-110K	AP	58.7	Cascade-RCNN

Related Papers

VoxDet: Rethinking 3D Semantic Occupancy Prediction as Dense Object Detection2025-06-05 Dense Object Detection Based on De-homogenized Queries2025-02-11 Hybrid Classification-Regression Adaptive Loss for Dense Object Detection2024-08-30 Sparse Generation: Making Pseudo Labels Sparse for Point Weakly Supervised Object Detection on Low Data Volume2024-03-28 Salience DETR: Enhancing Detection Transformer with Hierarchical Salience Filtering Refinement2024-03-24 Bridging Cross-task Protocol Inconsistency for Distillation in Dense Object Detection2023-08-28 CrossKD: Cross-Head Knowledge Distillation for Object Detection2023-06-20 Ambiguity-Resistant Semi-Supervised Learning for Dense Object Detection2023-03-27