# Anchor Loss: Modulating Loss Scale based on Prediction Difficulty

arXiv: https://arxiv.org/abs/1909.11155 · HTML: https://mlsanity.com/paper/1909.11155/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Sep 2019.

## Built on by (most cited first, 14 of 14)

- [Distributional Robustness Loss for Long-tail Learning](https://mlsanity.com/paper/2104.03066/) (2021)
- [Orthogonal Projection Loss](https://mlsanity.com/paper/2103.14021/) (2021)
- [Feedback-guided Data Synthesis for Imbalanced Classification](https://mlsanity.com/paper/2310.00158/) (2023)
- [Self-Supervised Keypoint Discovery in Behavioral Videos](https://mlsanity.com/paper/2112.05121/) (2021)
- [Simplifying Neural Network Training Under Class Imbalance](https://mlsanity.com/paper/2312.02517/) (2023)
- [Unsupervised Hyperbolic Representation Learning via Message Passing Auto-Encoders](https://mlsanity.com/paper/2103.16046/) (2021)
- [SYNAuG: Exploiting Synthetic Data for Data Imbalance Problems](https://mlsanity.com/paper/2308.00994/) (2023)
- [An Adversarial Human Pose Estimation Network Injected with Graph Structure](https://mlsanity.com/paper/2103.15534/) (2021)
- [Uncertainty-guided Boundary Learning for Imbalanced Social Event Detection](https://mlsanity.com/paper/2310.19247/) (2023)
- [Joint Learning of Blind Super-Resolution and Crack Segmentation for Realistic Degraded Images](https://mlsanity.com/paper/2302.12491/) (2023)
- [Weakly Supervised Keypoint Discovery](https://mlsanity.com/paper/2109.13423/) (2021)
- [Automatic lobe segmentation using attentive cross entropy and end-to-end fissure generation](https://mlsanity.com/paper/2307.12634/) (2023)
- [Streaming Anchor Loss: Augmenting Supervision with Temporal Significance](https://mlsanity.com/paper/2310.05886/) (2023)
- [Learning from Weakly-labeled Web Videos via Exploring Sub-Concepts](https://mlsanity.com/paper/2101.03713/) (2021)

## Abstract (arXiv)

We propose a novel loss function that dynamically rescales the cross entropy based on prediction difficulty regarding a sample. Deep neural network architectures in image classification tasks struggle to disambiguate visually similar objects. Likewise, in human pose estimation symmetric body parts often confuse the network with assigning indiscriminative scores to them. This is due to the output prediction, in which only the highest confidence label is selected without taking into consideration a measure of uncertainty. In this work, we define the prediction difficulty as a relative property coming from the confidence score gap between positive and negative labels. More precisely, the proposed loss function penalizes the network to avoid the score of a false prediction being significant. To demonstrate the efficacy of our loss function, we evaluate it on two different domains: image classification and human pose estimation. We find improvements in both applications by achieving higher accuracy compared to the baseline methods.