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.
Deep Context Maps: Agent Trajectory Prediction using Location-specific Latent Maps
The Garden of Forking Paths: Towards Multi-Future Trajectory Prediction
Totally Deep Support Vector Machines
HalluciNet-ing Spatiotemporal Representations Using a 2D-CNN
Flow-Distilled IP Two-Stream Networks for Compressed Video Action Recognition
SoccerDB: A Large-Scale Database for Comprehensive Video Understanding
Bi-Semantic Reconstructing Generative Network for Zero-shot Learning
Synthetic Humans for Action Recognition from Unseen Viewpoints
DCIL: Deep Contextual Internal Learning for Image Restoration and Image Retargeting
Self-supervised Object Motion and Depth Estimation from Video
Video action detection by learning graph-based spatio-temporal interactions
Robust, Extensible, and Fast: Teamed Classifiers for Vehicle Tracking and Vehicle Re-ID in Multi-Camera Networks
Learning Sparse 2D Temporal Adjacent Networks for Temporal Action Localization
DASZL: Dynamic Action Signatures for Zero-shot Learning
View-invariant Deep Architecture for Human Action Recognition using late fusion
Spatio-Temporal Pyramid Graph Convolutions for Human Action Recognition and Postural Assessment
RSA: Randomized Simulation as Augmentation for Robust Human Action Recognition
Skeleton based Activity Recognition by Fusing Part-wise Spatio-temporal and Attention Driven Residues
More Is Less: Learning Efficient Video Representations by Big-Little Network and Depthwise Temporal Aggregation
View-Invariant Probabilistic Embedding for Human Pose
Sparse and Low-Rank High-Order Tensor Regression via Parallel Proximal Method
Learning Structured Representations of Spatial and Interactive Dynamics for Trajectory Prediction in Crowded Scenes
Action Recognition via Pose-Based Graph Convolutional Networks with Intermediate Dense Supervision
DiversityGAN: Diversity-Aware Vehicle Motion Prediction via Latent Semantic Sampling
PREDICT & CLUSTER: Unsupervised Skeleton Based Action Recognition
An Attention-Enhanced Recurrent Graph Convolutional Network for Skeleton-Based Action Recognition
SRG: Snippet Relatedness-based Temporal Action Proposal Generator
Skeleton based Zero Shot Action Recognition in Joint Pose-Language Semantic Space
G-TAD: Sub-Graph Localization for Temporal Action Detection
Revisiting Deep Architectures for Head Motion Prediction in 360° Videos
Background Suppression Network for Weakly-supervised Temporal Action Localization
Incremental Learning of Motion Primitives for Pedestrian Trajectory Prediction at Intersections
Heterogeneous Graph-based Knowledge Transfer for Generalized Zero-shot Learning
Constrained R-CNN: A general image manipulation detection model
Deep Anomaly Detection with Deviation Networks
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.