Image Annotation in Presence of Noisy Labels
Source
Evernote/IFTTT Feedly/Image Annotation in Presence of Noisy Labels.md
Summary
이 논문은 소셜 이미지에서 발생하는 노이즈가 많은 라벨(사용자 협업 태그 등)을 활용하여 강건한 이미지 주석을 생성하는 프레임워크를 제안합니다. 노이즈가 있는 라벨의 공현(co-occurrence) 쌍에 내재된 집단 지식을 활용하여 기존 알고리즘 대비 Corel5K 및 ESP 데이터셋에서 더 높은 주석 정확도를 달성함을 입증했습니다.
Key Points
- 소셜 이미지 라벨의 노이즈 문제(사용자 협업 태그 등) 해결을 위한 새로운 프레임워크 제안
- 노이즈가 있는 라벨의 공현 관계(co-occurrence pairs)에서 집단 지식을 추출하여 강건한 주석 도출
- 기존 알고리즘 대비 Corel5K, ESP 벤치마크 데이터셋에서 주석 정확도 우월성 입증
Related
-
Protecting Sensitive Labels in Social Network Data Anonymization
-
웹 데이터베이스 검색 결과 자동 주석 처리 (Automatic Annotation of Web Database Search Results)
-
Moment-Based Spectral Analysis of Large-Scale Networks Using Local Structural Information
-
Continuous Birdsong Recognition Using Gaussian Mixture Modeling of Image Shape Features
-
Efficient Multiview Maintenance under Insertion in Huge Social Networks
-
Anomaly Extraction in Backbone Networks Using Association Rules
-
동적 스타 네트워크에서 다중 유형 객체의 공진화 (Co-Evolution of Multi-Typed Objects in Dynamic Star Networks)
-
Weakly Supervised Learning of Object Segmentations from Web-Scale Video
-
An Unsupervised Feature Selection Framework for Social Media Data
-
Social Event Classification via Boosted Multimodal Supervised Latent Dirichlet Allocation
-
Information-Theoretic Outlier Detection for Large-Scale Categorical Data
-
Active Learning through Adaptive Heterogeneous Ensembling (AHE)
-
Feature Ensemble Plus Sample Selection: Domain Adaptation for Sentiment Classification
-
A Model for Context in the Design of Open Production Communities
-
Efficient Estimation of Word Representations in Vector Space
-
Efficient Closed-Form Solution to Generalized Boundary Detection
-
Fast Near-Duplicate Image Detection Using Uniform Randomized Trees
-
Fast, Accurate Detection of 100,000 Object Classes on a Single Machine (Technical Supplement)
-
A Hamming Embedding Kernel with Informative Bag-of-Visual Words for Video Semantic Indexing
-
People reidentification in surveillance and forensics: A survey
-
Unsupervised Spatial Event Detection in Targeted Domains with Applications to Civil Unrest Modeling
-
Structured Streaming Skeleton (SSS): 온라인 인간 제스처 인식용 새로운 특징 추출 방법
-
Enlisting the Ghost: Modeling Empty Categories for Machine Translation
-
Smooth Nonnegative Matrix Factorization for Unsupervised Audiovisual Document Structuring
-
언어 독립적 시간 표현 판별적 파싱 (Language-Independent Discriminative Parsing of Temporal Expressions)
-
이동성 모델 분류 체계 및 조사 (A Taxonomy and Survey of Microscopic Mobility Models)
-
웹 페이지의 시각적 복잡성 측정 (Measuring the Visual Complexities of Web Pages)
-
Similarity-based Clustering by Left-Stochastic Matrix Factorization
-
Supporting Flexible, Efficient, and User-Interpretable Retrieval of Similar Time Series
-
Improved Domain Adaptation for Statistical Machine Translation
-
When Amazon Meets Google: Product Visualization by Exploring Multiple Web Sources
-
Online Graph Edge-Coloring in the Random-Order Arrival Model
-
Joint consideration of energy-efficiency and coverage-preservation in microsensor networks
-
Neighborhood Preserving Codes for Assigning Point Labels: Applications to Stochastic Search
-
Efficient Inference and Structured Learning for Semantic Role Labeling
-
Point Representation for Local Optimization: Towards Multi-Dimensional Gray Codes