MLTU: mixup long-tail unsupervised zero-shot image classification on vision-language models.
Yunpeng JiaXiufen YeXinkui MeiYusong LiuShuxiang GuoPublished in: Multim. Syst. (2024)
Keyphrases
- language model
- long tail
- image classification
- language modeling
- document retrieval
- speech recognition
- computer vision
- n gram
- language modelling
- probabilistic model
- bag of words
- information retrieval
- retrieval model
- image representation
- test collection
- statistical language models
- feature extraction
- image features
- query expansion
- supervised learning
- smoothing methods
- recommendation systems
- query terms
- semi supervised
- multi label
- social media
- online advertising
- visual words
- relevance model
- web search
- information extraction
- end users
- document ranking
- search engine
- data mining