# Localization Uncertainty Estimation for Anchor-Free Object Detection

arXiv: https://arxiv.org/abs/2006.15607 · HTML: https://mlsanity.com/paper/2006.15607/

## Key facts

- Cited by 19 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 28 Jun 2020.

## Built on by (most cited first, 19 of 19)

- [Human Pose Regression with Residual Log-likelihood Estimation](https://mlsanity.com/paper/2107.11291/) (2021)
- [Unbiased Teacher v2: Semi-supervised Object Detection for Anchor-free and Anchor-based Detectors](https://mlsanity.com/paper/2206.09500/) (2022)
- [Few-shot Object Counting and Detection](https://mlsanity.com/paper/2207.10988/) (2022)
- [The Multi-modality Cell Segmentation Challenge: Towards Universal Solutions](https://mlsanity.com/paper/2308.05864/) (2023)
- [Estimating and Evaluating Regression Predictive Uncertainty in Deep Object Detectors](https://mlsanity.com/paper/2101.05036/) (2021)
- [Uncertainty for Identifying Open-Set Errors in Visual Object Detection](https://mlsanity.com/paper/2104.01328/) (2021)
- [Adaptive Bounding Box Uncertainties via Two-Step Conformal Prediction](https://mlsanity.com/paper/2403.07263/) (2024)
- [iFS-RCNN: An Incremental Few-shot Instance Segmenter](https://mlsanity.com/paper/2205.15562/) (2022)
- [Gradient-Based Quantification of Epistemic Uncertainty for Deep Object Detectors](https://mlsanity.com/paper/2107.04517/) (2021)
- [CertainNet: Sampling-free Uncertainty Estimation for Object Detection](https://mlsanity.com/paper/2110.01604/) (2021)
- [Semi-Supervised Teacher-Reference-Student Architecture for Action Quality Assessment](https://mlsanity.com/paper/2407.19675/) (2024)
- [ACT-Net: Anchor-context Action Detection in Surgery Videos](https://mlsanity.com/paper/2310.03377/) (2023)
- [UAHOI: Uncertainty-aware Robust Interaction Learning for HOI Detection](https://mlsanity.com/paper/2408.07430/) (2024)
- [Object Detection as Probabilistic Set Prediction](https://mlsanity.com/paper/2203.07980/) (2022)
- [Uncertainty-Aware Gradient Stabilization for Small Object Detection](https://mlsanity.com/paper/2303.01803/) (2023)
- [Informative Data Selection with Uncertainty for Multi-modal Object Detection](https://mlsanity.com/paper/2304.11697/) (2023)
- [RPT++: Customized Feature Representation for Siamese Visual Tracking](https://mlsanity.com/paper/2110.12194/) (2021)
- [UTrack: Multi-Object Tracking with Uncertain Detections](https://mlsanity.com/paper/2408.17098/) (2024)
- [Rethinking Detection Calibration: A Coordinate and Direction Perspective](https://mlsanity.com/paper/2607.29040/) (2026)

## Abstract (arXiv)

Since many safety-critical systems, such as surgical robots and autonomous driving cars operate in unstable environments with sensor noise and incomplete data, it is desirable for object detectors to take the localization uncertainty into account. However, there are several limitations of the existing uncertainty estimation methods for anchor-based object detection. 1) They model the uncertainty of the heterogeneous object properties with different characteristics and scales, such as location (center point) and scale (width, height), which could be difficult to estimate. 2) They model box offsets as Gaussian distributions, which is not compatible with the ground truth bounding boxes that follow the Dirac delta distribution. 3) Since anchor-based methods are sensitive to anchor hyper-parameters, their localization uncertainty could also be highly sensitive to the choice of hyper-parameters. To tackle these limitations, we propose a new localization uncertainty estimation method called UAD for anchor-free object detection. Our method captures the uncertainty in four directions of box offsets (left, right, top, bottom) that are homogeneous, so that it can tell which direction is uncertain, and provide a quantitative value of uncertainty in [0, 1]. To enable such uncertainty estimation, we design a new uncertainty loss, negative power log-likelihood loss, to measure the localization uncertainty by weighting the likelihood loss by its IoU, which alleviates the model misspecification problem. Furthermore, we propose an uncertainty-aware focal loss for reflecting the estimated uncertainty to the classification score. Experimental results on COCO datasets demonstrate that our method significantly improves FCOS, by up to 1.8 points, without sacrificing computational efficiency.