PP-OCRv6 released on Hugging Face, claims to surpass billion-scale VLMs

Arxiv CS.CV (computer vision)

PP-OCRv6 uses 34.5M parameters to outperform billion-parameter VLMs on OCR tasks, according to a new paper. PaddleOCR 3.7 also adds transformers and ONNX Runtime backends.

PaddleOCRHugging FaceOCR

See 2 more sources

RT @slimcat0101: PP-OCRv6 is now on @HuggingFace! 🎉 Not just better accuracy— PaddleOCR 3.7 also...5 days agoHugging Face

PaddleOCR-VL-1.6: Expanding the Frontier of Document Parsing with Under-Optimized Region Refinement and Progressive Post-Training28 days agoZelun Zhang, Hongen Liu, Suyin Liang, Yubo Zhang, Yiqing Xiang, Jiaxuan Liu, Ting Sun, Manhui Lin, Yue Zhang, Changda Zhou, Tingquan Gao, Cheng Cui, Yi Liu, Dianhai Yu, Yanjun Ma

Arxiv CS.CV (computer vision)·Yubo ZhangXueqing WangManhui LinYue ZhangPenglongyi DengTing SunTingquan GaoZelun ZhangJiaxuan LiuChangda ZhouHongen LiuSuyin LiangCheng CuiYi LiuDianhai YuYanjun Ma

Jun 12, 4:00 AM