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.
Improving Automated Program Repair with Domain Adaptation
On Improving Summarization Factual Consistency from Natural Language Feedback
Empowering Sentence Encoders with Prompting and Label Retrieval for Zero-shot Text Classification
Go-tuning: Improving Zero-shot Learning Abilities of Smaller Language Models
Precise Zero-Shot Dense Retrieval without Relevance Labels
A Survey on Human Action Recognition
Distilling Vision-Language Pre-training to Collaborate with Weakly-Supervised Temporal Action Localization
BLOOM+1: Adding Language Support to BLOOM for Zero-Shot Prompting
LaSQuE: Improved Zero-Shot Classification from Explanations Through Quantifier Modeling and Curriculum Learning
iCub! Do you recognize what I am doing?: multimodal human action recognition on multisensory-enabled iCub robot
Attentive Mask CLIP
Pre-trained Language Models Can be Fully Zero-Shot Learners
Significantly improving zero-shot X-ray pathology classification via fine-tuning pre-trained image-text encoders
NLIP: Noise-robust Language-Image Pre-training
Reproducible scaling laws for contrastive language-image learning
Dilation-Erosion for Single-Frame Supervised Temporal Action Localization
LidarCLIP or: How I Learned to Talk to Point Clouds
Resolving Semantic Confusions for Improved Zero-Shot Detection
Cap2Aug: Caption guided Image to Image data Augmentation
FLAG3D: A 3D Fitness Activity Dataset with Language Instruction
Leveraging Spatio-Temporal Dependency for Skeleton-Based Action Recognition
Reliable Multimodal Trajectory Prediction via Error Aligned Uncertainty Optimization
Open Vocabulary Semantic Segmentation with Patch Aligned Contrastive Learning
A Temporal Graph Neural Network for Cyber Attack Detection and Localization in Smart Grids
ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation
Towards Explainable Motion Prediction using Heterogeneous Graph Representations
X-Paste: Revisiting Scalable Copy-Paste for Instance Segmentation using CLIP and StableDiffusion
InternVideo: General Video Foundation Models via Generative and Discriminative Learning
Hierarchical Contrast for Unsupervised Skeleton-based Action Representation Learning
Perceive, Interact, Predict: Learning Dynamic and Static Clues for End-to-End Motion Prediction
Applying Multilingual Models to Question Answering (QA)
ResFormer: Scaling ViTs with Multi-Resolution Training
Random Copolymer inverse design system orienting on Accurate discovering of Antimicrobial peptide-mimetic copolymers
Zero-Shot Learning for Joint Intent and Slot Labeling
SuS-X: Training-Free Name-Only Transfer of Vision-Language Models
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.