# Inter-Class Relational Loss for Small Object Detection: A Case Study on License Plates

arXiv: https://arxiv.org/abs/2508.14343 · HTML: https://mlsanity.com/paper/2508.14343/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 36 references that are in the map.
- First posted on arXiv 20 Aug 2025.

## Builds on (most cited first, 22 of 22)

- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Generalized Intersection over Union: A Metric and A Loss for Bounding Box Regression](https://mlsanity.com/paper/1902.09630/) (2019)
- [On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima](https://mlsanity.com/paper/1609.04836/) (2016)
- [DETRs Beat YOLOs on Real-time Object Detection](https://mlsanity.com/paper/2304.08069/) (2023)
- [Distance-IoU Loss: Faster and Better Learning for Bounding Box Regression](https://mlsanity.com/paper/1911.08287/) (2019)
- [YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information](https://mlsanity.com/paper/2402.13616/) (2024)
- [YOLOv11: An Overview of the Key Architectural Enhancements](https://mlsanity.com/paper/2410.17725/) (2024)
- [YOLOv12: Attention-Centric Real-Time Object Detectors](https://mlsanity.com/paper/2502.12524/) (2025)
- [IoU Loss for 2D/3D Object Detection](https://mlsanity.com/paper/1908.03851/) (2019)
- [Towards Large-Scale Small Object Detection: Survey and Benchmarks](https://mlsanity.com/paper/2207.14096/) (2022)
- [Enhancing Geometric Factors in Model Learning and Inference for Object Detection and Instance Segmentation](https://mlsanity.com/paper/2005.03572/) (2020)
- [RT-DETRv2: Improved Baseline with Bag-of-Freebies for Real-Time Detection Transformer](https://mlsanity.com/paper/2407.17140/) (2024)
- [Scale Match for Tiny Person Detection](https://mlsanity.com/paper/1912.10664/) (2019)
- [A Robust Real-Time Automatic License Plate Recognition Based on the YOLO Detector](https://mlsanity.com/paper/1802.09567/) (2018)
- [Factors in Finetuning Deep Model for object detection](https://mlsanity.com/paper/1601.05150/) (2016)
- [A Normalized Gaussian Wasserstein Distance for Tiny Object Detection](https://mlsanity.com/paper/2110.13389/) (2021)
- [DQ-DETR: DETR with Dynamic Query for Tiny Object Detection](https://mlsanity.com/paper/2404.03507/) (2024)
- [Small Object Detection via Coarse-to-fine Proposal Generation and Imitation Learning](https://mlsanity.com/paper/2308.09534/) (2023)
- [IoU-balanced Loss Functions for Single-stage Object Detection](https://mlsanity.com/paper/1908.05641/) (2019)
- [UAV-DETR: Efficient End-to-End Object Detection for Unmanned Aerial Vehicle Imagery](https://mlsanity.com/paper/2501.01855/) (2025)
- [Unified Chinese License Plate Detection and Recognition with High Efficiency](https://mlsanity.com/paper/2205.03582/) (2022)
- [Focaler-IoU: More Focused Intersection over Union Loss](https://mlsanity.com/paper/2401.10525/) (2024)

## Abstract (arXiv)

In one-stage multi-object detection tasks, various intersection over union (IoU)-based solutions aim at smooth and stable convergence near the targets during training. However, IoU-based losses fail to correctly update the gradient of small objects due to an extremely flat gradient. During the update of multiple objects, the learning of small objects' gradients suffers more because of insufficient gradient updates. Therefore, we propose an inter-class relational loss to efficiently update the gradient of small objects while not sacrificing the learning efficiency of other objects based on the simple fact that an object has a spatial relationship to another object (e.g., a car plate is attached to a car in a similar position). When the predicted car plate's bounding box is not within its car, a loss punishment is added to guide the learning, which is inversely proportional to the overlapped area of the car's and predicted car plate's bounding box. By leveraging the spatial relationship at the inter-class level, the loss guides small object predictions using larger objects and enhances latent information in deeper feature maps. In this paper, we present twofold contributions using license plate detection as a case study: (1) a new small vehicle multi-license plate dataset (SVMLP), featuring diverse real-world scenarios with high-quality annotations; and (2) a novel inter-class relational loss function designed to promote effective detection performance. We highlight the proposed ICR loss penalty can be easily added to existing IoU-based losses and enhance the performance. These contributions improve the standard mean Average Precision (mAP) metric, achieving gains of 10.3% and 1.6% in mAPtest50 for YOLOv12-T and UAV-DETR, respectively, without any additional hyperparameter tuning. Code and dataset will be available soon.