NVlabs/HMAR
[CVPR 2025] HMAR: Efficient Hierarchical Masked Auto-Regressive Image Generation
This project helps researchers and machine learning engineers working on computer vision tasks to efficiently generate high-quality images. You input a large dataset of images (like ImageNet) and specify the desired image resolution, then the system produces new, diverse images that match the characteristics of your training data. This is ideal for those focused on developing or evaluating new generative image models.
No commits in the last 6 months.
Use this if you need to generate high-resolution, class-conditional images from a given dataset for research or model evaluation purposes.
Not ideal if you are looking for a pre-trained model for immediate use in a consumer application or if you don't have access to substantial computational resources for training.
Stars
61
Forks
2
Language
Python
License
—
Category
Last pushed
Jul 08, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/transformers/NVlabs/HMAR"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.