DeTiME: Diffusion-Enhanced Topic Modeling using Encoder-decoder based LLM.
Weijie XuWenxiang HuFanyou WuSrinivasan H. SengameduPublished in: EMNLP (Findings) (2023)
Keyphrases
- topic modeling
- video codec
- decoding process
- topic models
- distributed video coding
- low complexity
- error control
- noisy channel
- wyner ziv video coding
- text mining
- wyner ziv
- modeling framework
- successive approximation
- latent dirichlet allocation
- text classification
- distributed source coding
- bit rate
- rate distortion
- error resilience
- video coding
- turbo codes
- video compression
- motion estimation
- text corpora
- latent topics
- transform domain
- error concealment
- collaborative filtering
- information retrieval
- real world
- error correction
- pairwise
- computational complexity
- decoding algorithm
- image quality
- natural language processing
- bit budget
- artificial intelligence