Zero-Shot Learning
태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.
- 하위 태스크
- 7
- 평가표
- 33
- 보존된 평가 기록
- 84
Zero-shot learning (ZSL) is a model's ability to detect classes never seen during training. The condition is that the classes are not known during supervised learning.
Earlier work in zero-shot learning use attributes in a two-step approach to infer unknown classes. In the computer vision context, more recent advances learn mappings from image feature space to semantic space. Other approaches learn non-linear multimodal embeddings. In the modern NLP context, language models can be evaluated on downstream tasks without fine tuning.
Benchmark datasets for zero-shot learning include aPY, AwA, and CUB, among others.
( Image credit: Prototypical Networks for Few shot Learning in PyTorch )
Further readings:
논문
이 태스크와 하위 분야의 관련 논문입니다. 실제 성능 비교는 평가 결과에서 확인하세요.
Mining Mid-level Features for Action Recognition Based on Effective Skeleton Representation
10,000+ Times Accelerated Robust Subset Selection (ARSS)
Very Deep Convolutional Networks for Large-Scale Image Recognition
Action Recognition in the Frequency Domain
Sparse Coding on Symmetric Positive Definite Manifolds using Bregman Divergences
Temporal Extension of Scale Pyramid and Spatial Pyramid Matching for Action Recognition
Down-Sampling coupled to Elastic Kernel Machines for Efficient Recognition of Isolated Gestures
HOPC: Histogram of Oriented Principal Components of 3D Pointclouds for Action Recognition
Robust Statistical Approach for Extraction of Moving Human Silhouettes from Videos
Multithreshold Entropy Linear Classifier
A discussion on the validation tests employed to compare human action recognition methods using the MSR Action3D dataset
Feature and Region Selection for Visual Learning
Early Recognition of Human Activities from First-Person Videos Using Onset Representations
Semantic Graph for Zero-Shot Learning
Log-Euclidean Bag of Words for Human Action Recognition
Two-Stream Convolutional Networks for Action Recognition in Videos
Robust Estimation of 3D Human Poses from a Single Image
Continuous Action Recognition Based on Sequence Alignment
Feature sampling and partitioning for visual vocabulary generation on large action classification datasets
Bag of Visual Words and Fusion Methods for Action Recognition: Comprehensive Study and Good Practice
Cross-view Action Modeling, Learning and Recognition
Shrinkage Optimized Directed Information using Pictorial Structures for Action Recognition
KPCA Spatio-temporal trajectory point cloud classifier for recognizing human actions in a CBVR system
Bregman Divergences for Infinite Dimensional Covariance Matrices
Multiview Hessian regularized logistic regression for action recognition
A Hybrid Loss for Multiclass and Structured Prediction
Extrinsic Methods for Coding and Dictionary Learning on Grassmann Manifolds
Semantic Similarity Measures Applied to an Ontology for Human-Like Interaction
Actions in the Eye: Dynamic Gaze Datasets and Learnt Saliency Models for Visual Recognition
EXMOVES: Classifier-based Features for Scalable Action Recognition
Zero-Shot Learning for Semantic Utterance Classification
Zero-Shot Learning by Convex Combination of Semantic Embeddings
Dictionary Learning and Sparse Coding on Grassmann Manifolds: An Extrinsic Solution
Combining Spatio-Temporal Appearance Descriptors and Optical Flow for Human Action Recognition in Video Data
Seeing What You're Told: Sentence-Guided Activity Recognition In Video
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.