paper-with-me

Papers

De-biasing Weakly Supervised Learning by Regularizing Prediction Entropy

2019-03-20 · ICLR Workshop LLD 2019 · Dean Wyatte

We explore the effect of regularizing prediction entropy in a weakly supervised setting with inexact class labels. When underlying data distributions are biased toward a specific subclass, we hypothesize that entropy regularization can be used to bootstrap a training set that mitigates this bias. We conduct experiments over multiple datasets under supervision of an oracle and in a semi-supervised setting finding substantial reductions in training set bias capable of decreasing test error rate. These findings suggest entropy regularization as a promising approach to de-biasing weakly supervised learning systems.

📄 PDF Abstract BibTeX

Code (0)

등록된 구현이 없습니다.

Tasks

Weakly-supervised Learning

Similar Papers 제목 키워드 기반

Debate-Enhanced Pseudo Labeling and Frequency-Aware Progressive Debiasing for Weakly-Supervised Camouflaged Object Detection with Scribble Annotations

2025-12-23 · Jiawei Ge, Jiuxin Cao, Xinyi Li, Xuelin Zhu 외 arxiv

Weakly-Supervised Camouflaged Object Detection (WSCOD) aims to locate and segment objects that are visually concealed within their surrounding scenes, relying solely on sparse supervision such as scribble annotations. De…

Object Detection

Adaptation of Weakly Supervised Localization in Histopathology by Debiasing Predictions

2026-03-12 · Alexis Guichemerre, Banafsheh Karimian, Soufiane Belharbi, Natacha Gillet 외 arxiv

Weakly Supervised Object Localization (WSOL) models enable joint classification and region-of-interest localization in histology images using only image-class supervision. When deployed in a target domain, distributions …

Object LocalizationDomain Adaptation

Using Focal Loss to Fight Shallow Heuristics: An Empirical Analysis of Modulated Cross-Entropy in Natural Language Inference

2022-11-23 · Frano Rajič, Ivan Stresec, Axel Marmet, Tim Poštuvan

There is no such thing as a perfect dataset. In some datasets, deep neural networks discover underlying heuristics that allow them to take shortcuts in the learning process, resulting in poor generalization capability. I…

Natural Language Inference

Information Entropy Based Feature Pooling for Convolutional Neural Networks

2019-10-01 · ICCV 2019 10 · Weitao Wan, Jiansheng Chen, Tianpeng Li, Yiqing Huang 외

In convolutional neural networks (CNNs), we propose to estimate the importance of a feature vector at a spatial location in the feature maps by the network's uncertainty on its class prediction, which can be quantified u…

General ClassificationObject LocalizationSegmentationSemantic Segmentation+3

Min-Entropy Latent Model for Weakly Supervised Object Detection

2019-02-16 · CVPR 2018 6 · Fang Wan, Pengxu Wei, Zhenjun Han, Jianbin Jiao 외

Weakly supervised object detection is a challenging task when provided with image category supervision but required to learn, at the same time, object locations and object detectors. The inconsistency between the weak su…

image-classificationImage ClassificationObjectobject-detection+4