Xray-Visual Models: Scaling Vision models on Industry Scale Data
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914338627911680 |
|---|---|
| author | Mishra, Shlok Lin, Tsung-Yu Wang, Linda Xu, Hongli Liu, Yimin Hsu, Michael Ahuja, Chaitanya Yuan, Hao Cheng, Jianpeng Chen, Hong-You Xu, Haoyuan Li, Chao Awasthi, Abhijeet Moon, Jihye Husa, Don Ge, Michael Singla, Sumedha Chowdhury, Arkabandhu Dingh, Phong Shukla, Satya Narayan Yang, Yonghuan Jacobs, David Guo, Qi Xiao, Jun Fan, Xiangjun Singh, Aashu |
| author_facet | Mishra, Shlok Lin, Tsung-Yu Wang, Linda Xu, Hongli Liu, Yimin Hsu, Michael Ahuja, Chaitanya Yuan, Hao Cheng, Jianpeng Chen, Hong-You Xu, Haoyuan Li, Chao Awasthi, Abhijeet Moon, Jihye Husa, Don Ge, Michael Singla, Sumedha Chowdhury, Arkabandhu Dingh, Phong Shukla, Satya Narayan Yang, Yonghuan Jacobs, David Guo, Qi Xiao, Jun Fan, Xiangjun Singh, Aashu |
| contents | We present Xray-Visual, a unified vision model architecture for large-scale image and video understanding trained on industry-scale social media data. Our model leverages over 15 billion curated image-text pairs and 10 billion video-hashtag pairs from Facebook and Instagram, employing robust data curation pipelines that incorporate balancing and noise suppression strategies to maximize semantic diversity while minimizing label noise. We introduce a three-stage training pipeline that combines self-supervised MAE, semi-supervised hashtag classification, and CLIP-style contrastive learning to jointly optimize image and video modalities. Our architecture builds on a Vision Transformer backbone enhanced with efficient token reorganization (EViT) for improved computational efficiency. Extensive experiments demonstrate that Xray-Visual achieves state-of-the-art performance across diverse benchmarks, including ImageNet for image classification, Kinetics and HMDB51 for video understanding, and MSCOCO for cross-modal retrieval. The model exhibits strong robustness to domain shift and adversarial perturbations. We further demonstrate that integrating large language models as text encoders (LLM2CLIP) significantly enhances retrieval performance and generalization capabilities, particularly in real-world environments. Xray-Visual establishes new benchmarks for scalable, multimodal vision models, while maintaining superior accuracy and computational efficiency. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_16918 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Xray-Visual Models: Scaling Vision models on Industry Scale Data Mishra, Shlok Lin, Tsung-Yu Wang, Linda Xu, Hongli Liu, Yimin Hsu, Michael Ahuja, Chaitanya Yuan, Hao Cheng, Jianpeng Chen, Hong-You Xu, Haoyuan Li, Chao Awasthi, Abhijeet Moon, Jihye Husa, Don Ge, Michael Singla, Sumedha Chowdhury, Arkabandhu Dingh, Phong Shukla, Satya Narayan Yang, Yonghuan Jacobs, David Guo, Qi Xiao, Jun Fan, Xiangjun Singh, Aashu Computer Vision and Pattern Recognition Artificial Intelligence We present Xray-Visual, a unified vision model architecture for large-scale image and video understanding trained on industry-scale social media data. Our model leverages over 15 billion curated image-text pairs and 10 billion video-hashtag pairs from Facebook and Instagram, employing robust data curation pipelines that incorporate balancing and noise suppression strategies to maximize semantic diversity while minimizing label noise. We introduce a three-stage training pipeline that combines self-supervised MAE, semi-supervised hashtag classification, and CLIP-style contrastive learning to jointly optimize image and video modalities. Our architecture builds on a Vision Transformer backbone enhanced with efficient token reorganization (EViT) for improved computational efficiency. Extensive experiments demonstrate that Xray-Visual achieves state-of-the-art performance across diverse benchmarks, including ImageNet for image classification, Kinetics and HMDB51 for video understanding, and MSCOCO for cross-modal retrieval. The model exhibits strong robustness to domain shift and adversarial perturbations. We further demonstrate that integrating large language models as text encoders (LLM2CLIP) significantly enhances retrieval performance and generalization capabilities, particularly in real-world environments. Xray-Visual establishes new benchmarks for scalable, multimodal vision models, while maintaining superior accuracy and computational efficiency. |
| title | Xray-Visual Models: Scaling Vision models on Industry Scale Data |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2602.16918 |