Abstract:Webly supervised fine-grained recognition aims to distinguish subordinate categories (e.g., bird species) with freely available web data. It has significant research and application value for alleviating the costly professional manual annotations' dependence in the fine-grained recognition task. Nevertheless, there exists label noise in web data to decrease the model's recognition performance. Most existing methods attempt to select clean data via loss analyses, which favors easy samples to hinder mining subtle differences contained in hard samples. Inspired by the intrinsic trait of consistent semantic predictions among different hierarchies of clean samples in fine-grained recognition, we propose a hierarchical consistency learning (HCL) approach for detecting noisy samples and capturing multi-hierarchy discriminative clues simultaneously. Specifically, our HCL approach works in a coarse-to-fine order, which first explores the semantic consistency between the image level and object level through prediction distribution conformance analyses. The open-set noise (i.e., samples irrelevant to any fine-grained subcategory) is thus detected, and the visual object information is highlighted with image-object contrastive learning. Then, the semantic consistency between object-level and part-level prediction distributions is utilized for detecting closed-set noise (i.e., samples mislabeled as other fine-grained subcategories), and local discriminative information is enhanced with object-part contrastive learning. Extensive experiments and analyses on three widely-used webly supervised fine-grained benchmark datasets demonstrate that the proposed HCL approach can achieve new state-of-the-art. The code is available at https://github.com/PKU-ICST-MIPL/HCL_TMM2023.

Hierarchical Semi-supervised Contrastive Learning for Contamination-Resistant Anomaly Detection

Learning Discrimination from Contaminated Data: Multi-Instance Learning for Unsupervised Anomaly Detection

Unilaterally Aggregated Contrastive Learning with Hierarchical Augmentation for Anomaly Detection

RoSAS: Deep Semi-Supervised Anomaly Detection with Contamination-Resilient Continuous Supervision

Masked Contrastive Learning for Anomaly Detection

Learning Discriminative Features for Semi-Supervised Anomaly Detection.

HCL-MTSAD: Hierarchical Contrastive Consistency Learning for Accurate Detection of Industrial Multivariate Time Series Anomalies

CL-Flow:Strengthening the Normalizing Flows by Contrastive Learning for Better Anomaly Detection

HCL: Hierarchical Consistency Learning for Webly Supervised Fine-Grained Recognition

UAC-AD: Unsupervised Adversarial Contrastive Learning for Anomaly Detection on Multi-modal Data in Microservice Systems

Elsa: Energy-based learning for semi-supervised anomaly detection

Multilevel Saliency-Guided Self-Supervised Learning for Image Anomaly Detection

Time-series Anomaly Detection via Contextual Discriminative Contrastive Learning

A Generic Machine Learning Framework for Fully-Unsupervised Anomaly Detection with Contaminated Data

Contrastive Attention for Video Anomaly Detection

Hyperbolic Self-supervised Contrastive Learning Based Network Anomaly Detection

Hierarchical Semantic Contrast for Scene-aware Video Anomaly Detection

Anomaly Heterogeneity Learning for Open-set Supervised Anomaly Detection

ESAD: End-to-end Semi-supervised Anomaly Detection

Catching Both Gray and Black Swans: Open-set Supervised Anomaly Detection

Abnormal Event Detection via Hypergraph Contrastive Learning