Autoregressive Models in Vision surveys recent progress in autoregressive models within computer vision. This repository compiles a curated list of research papers exploring the latest advancements in this field. It addresses the growing interest in modeling data sequentially for tasks like image and video generation and highlights the shift toward more sophisticated autoregressive techniques.
This repository offers a focused and up-to-date collection of survey papers on autoregressive models in vision. It is regularly updated with the latest research, providing a valuable resource for researchers and practitioners. The categorization by task (image, video, 3D, multimodal) enhances navigability and facilitates targeted exploration of specific areas.
- Image Generation: Comprehensive collection of papers covering unconditional and conditional image generation techniques using autoregressive models.
- Video Generation: Includes research related to unconditional and conditional video generation approaches.
- 3D Generation: Features papers exploring 3D content generation methods based on autoregressive models.
- Multimodal Generation: Covers research on models that combine visual data with other modalities.
- Evaluation Metrics: Includes work on metrics for evaluating autoregressive models in vision.
- Tutorial: Provides resources and guides related to using autoregressive models.
- Reasoning Alignment: Addresses the aspect of aligning autoregressive models with reasoning capabilities.
The repository is actively maintained and updated with newly published survey papers. It includes a timeline of additions and revisions, indicating ongoing development. The inclusion of a contribution guide suggests a community-driven approach to knowledge sharing and refinement.
This repository is valuable for researchers, students, and practitioners interested in autoregressive models in computer vision. It provides a centralized and curated collection of recent survey papers, facilitating understanding of the field's state-of-the-art, identifying key trends, and supporting further research and development.