# Unbiased Regression Loss for DETRs

Source: [Unbiased Regression Loss for DETRs](<https://www.sota2.com/research/paper/68fa75781e209131cea682ac>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T03:35:52.229Z

Authors: Edric, Ueta Daisuke, Kurokawa Yukimasa, Karlekar Jayashree, Sugiri Pranata

Published: 2024-10-30T02:05:18.000Z

## Abstract

In this paper, we introduce a novel unbiased regression loss for DETR-based detectors. The conventional $L\_{1}$ regression loss tends to bias towards larger boxes, as they disproportionately contribute more towards the overall loss compared to smaller boxes. Consequently, the detection performance for small objects suffers. To alleviate this bias, the proposed new unbiased loss, termed Sized $L\_{1}$ loss, normalizes the size of all boxes based on their individual width and height. Our experiments demonstrate consistent improvements in both fully-supervised and semi-supervised settings using the MS-COCO benchmark dataset.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2410.22638>)

[PDF](<https://arxiv.org/pdf/2410.22638.pdf>)