{"id":36127,"date":"2025-01-06T13:00:11","date_gmt":"2025-01-06T05:00:11","guid":{"rendered":"https:\/\/17aitech.com\/?p=36127"},"modified":"2025-01-06T13:00:11","modified_gmt":"2025-01-06T05:00:11","slug":"neurips2024-ocr-omni%e6%9d%a5%e4%ba%86%ef%bc%8c%e5%ad%97%e8%8a%82%e5%8d%8e%e5%b8%88%e6%8f%90%e5%87%ba%e7%bb%9f%e4%b8%80%e7%9a%84%e5%a4%9a%e6%a8%a1%e6%80%81%e6%96%87%e5%ad%97%e7%90%86%e8%a7%a3","status":"publish","type":"post","link":"https:\/\/17aitech.com\/?p=36127","title":{"rendered":"NeurIPS2024 | OCR-Omni\u6765\u4e86\uff0c\u5b57\u8282&amp;\u534e\u5e08\u63d0\u51fa\u7edf\u4e00\u7684\u591a\u6a21\u6001\u6587\u5b57\u7406\u89e3\u4e0e\u751f\u6210\u5927\u6a21\u578b"},"content":{"rendered":"<p>\u6587\u7ae0\u6765\u6e90\u4e8e\u4e92\u8054\u7f51:<a href=\"https:\/\/www.jiqizhixin.com\/articles\/2024-10-16\" target=\"_blank\">NeurIPS2024 | OCR-Omni\u6765\u4e86\uff0c\u5b57\u8282&amp;\u534e\u5e08\u63d0\u51fa\u7edf\u4e00\u7684\u591a\u6a21\u6001\u6587\u5b57\u7406\u89e3\u4e0e\u751f\u6210\u5927\u6a21\u578b<\/a><\/p>\n<p><strong>\u7814\u7a76\u80cc\u666f\u4e0e\u6311\u6218<\/strong><\/p>\n<p>\u5728<mark data-id=\"2d28aa9c-942d-471d-bd96-8bfefb7144e0\" data-type=\"concepts\">\u4eba\u5de5\u667a\u80fd<\/mark>\u9886\u57df\uff0c\u8d4b\u4e88\u673a\u5668\u7c7b\u4eba\u7684\u56fe\u50cf\u6587\u5b57\u611f\u77e5\u3001\u7406\u89e3\u3001\u7f16\u8f91\u548c\u751f\u6210\u80fd\u529b\u4e00\u76f4\u662f\u7814\u7a76\u70ed\u70b9\u3002\u76ee\u524d\uff0c\u89c6\u89c9\u6587\u5b57\u9886\u57df\u7684\u5927\u6a21\u578b\u7814\u7a76\u4e3b\u8981\u805a\u7126\u4e8e\u5355\u6a21\u6001\u751f\u6210\u4efb\u52a1\u3002\u5c3d\u7ba1\u8fd9\u4e9b\u6a21\u578b\u5728\u67d0\u4e9b\u4efb\u52a1\u4e0a\u5b9e\u73b0\u4e86\u7edf\u4e00\uff0c\u4f46\u5728 OCR \u9886\u57df\u7684\u591a\u6570\u4efb\u52a1\u4e0a\u4ecd\u96be\u4ee5\u8fbe\u6210\u5168\u9762\u6574\u5408\u3002<\/p>\n<p>\u4f8b\u5982\uff0cMonkey \u7b49\u89c6\u89c9\u8bed\u8a00\u6a21\u578b\uff08VLM\uff09\u64c5\u957f\u6587\u5b57\u68c0\u6d4b\u3001\u8bc6\u522b\u548c\u89c6\u89c9\u95ee\u7b54\uff08VQA\uff09\u7b49\u6587\u672c\u6a21\u6001\u751f\u6210\u4efb\u52a1\uff0c\u5374\u65e0\u6cd5\u80dc\u4efb\u6587\u5b57\u56fe\u50cf\u7684\u751f\u6210\u3001\u62b9\u9664\u548c\u7f16\u8f91\u7b49\u56fe\u50cf\u6a21\u6001\u751f\u6210\u4efb\u52a1\u3002\u53cd\u4e4b\uff0c\u4ee5 AnyText \u4e3a\u4ee3\u8868\u7684\u57fa\u4e8e\u6269\u6563\u6a21\u578b\u7684<mark data-id=\"ba1ea8a1-a705-4f89-8348-ec71bae840dc\" data-type=\"tech_tasks\">\u56fe\u50cf\u751f\u6210<\/mark>\u6a21\u578b\u5219\u4e13\u6ce8\u4e8e\u56fe\u50cf\u521b\u5efa\u3002\u56e0\u6b64\uff0cOCR \u9886\u57df\u4e9f\u9700\u4e00\u4e2a\u80fd\u591f\u7edf\u4e00\u591a\u6a21\u6001\u751f\u6210\u7684\u5927\u6a21\u578b\u3002<\/p>\n<p><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-d80cfc719b4c053dacd27bffbb8f7b8c.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-d80cfc719b4c053dacd27bffbb8f7b8c.png\"><\/a><\/p>\n<p><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-14127ab0aba041794eb5c5d5c053be63.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-14127ab0aba041794eb5c5d5c053be63.png\"><\/a><\/p>\n<p>\u8bba\u6587\u94fe\u63a5: https:\/\/arxiv.org\/abs\/2407.16364<\/p>\n<p>\u4ee3\u7801\u5f00\u6e90: https:\/\/github.com\/bytedance\/TextHarmony<\/p>\n<p><strong>\u5173\u952e\u95ee\u9898<\/strong><\/p>\n<p><strong>\u591a\u6a21\u6001\u751f\u6210\u7684\u5185\u5728\u77db\u76fe<\/strong><\/p>\n<p>\u7814\u7a76\u4eba\u5458\u53d1\u73b0\uff0c\u591a\u6a21\u6001\u751f\u6210\u5927\u6a21\u578b\u9762\u4e34\u89c6\u89c9\u4e0e\u8bed\u8a00\u6a21\u6001\u4e4b\u95f4\u7684\u56fa\u6709\u4e0d\u4e00\u81f4\u6027\uff0c\u8fd9\u5f80\u5f80\u5bfc\u81f4\u6a21\u578b\u6027\u80fd\u663e\u8457\u4e0b\u6ed1\u3002\u5982\u56fe\u6240\u793a\uff0c\u5728\u6587\u672c\u751f\u6210\u4efb\u52a1\u4e0a\uff0c\u591a\u6a21\u6001<mark data-id=\"d92a1cda-8e30-4199-9678-2d8c6c17d736\" data-type=\"tech_methods\">\u751f\u6210\u6a21\u578b<\/mark>\u76f8\u6bd4\u5355\u6a21\u6001<mark data-id=\"d92a1cda-8e30-4199-9678-2d8c6c17d736\" data-type=\"tech_methods\">\u751f\u6210\u6a21\u578b<\/mark>\u6548\u679c\u964d\u4f4e5%\uff0c\u5728<mark data-id=\"ba1ea8a1-a705-4f89-8348-ec71bae840dc\" data-type=\"tech_tasks\">\u56fe\u50cf\u751f\u6210<\/mark>\u4e0a\u964d\u4f4e\u4e868%\u3002\u4e3a\u5e94\u5bf9\u8fd9\u4e00\u6311\u6218\uff0c\u8fd1\u671f\u7684\u4e00\u4e9b\u7814\u7a76\u91c7\u7528\u4e86\u7279\u5b9a\u6a21\u6001\u7684\u76d1\u7763\u5fae\u8c03\uff0c\u4ece\u800c\u5206\u522b\u4f18\u5316\u6587\u5b57\u751f\u6210\u548c\u56fe\u7247\u751f\u6210\u7684\u6a21\u578b\u6743\u91cd\u3002\u7136\u800c\uff0c\u8fd9\u79cd\u65b9\u6cd5\u4e0e\u7edf\u4e00\u89c6\u89c9\u7406\u89e3\u4e0e\u751f\u6210\u7684\u521d\u8877\u76f8\u6096\u3002<\/p>\n<p>\u4e3a\u89e3\u51b3\u8fd9\u4e00\u96be\u9898\uff0c\u5b57\u8282\u8df3\u52a8\u4e0e\u534e\u4e1c\u5e08\u8303\u5927\u5b66\u7684\u8054\u5408\u7814\u7a76\u56e2\u961f\u63d0\u51fa\u4e86\u521b\u65b0\u6027\u7684\u591a\u6a21\u6001<mark data-id=\"d92a1cda-8e30-4199-9678-2d8c6c17d736\" data-type=\"tech_methods\">\u751f\u6210\u6a21\u578b<\/mark> \u2014\u2014TextHarmony\u3002\u8be5\u6a21\u578b\u4e0d\u4ec5\u7cbe\u901a\u89c6\u89c9\u6587\u672c\u7684\u611f\u77e5\u3001\u7406\u89e3\u548c\u751f\u6210\uff0c\u8fd8\u5728\u5355\u4e00\u6a21\u578b\u67b6\u6784\u4e2d\u5b9e\u73b0\u4e86\u89c6\u89c9\u4e0e\u8bed\u8a00\u6a21\u6001\u751f\u6210\u7684\u548c\u8c10\u7edf\u4e00\u3002<\/p>\n<p><strong>TextHarmony: \u7a81\u7834\u6027\u8d21\u732e<\/strong><\/p>\n<p>TextHarmony \u7684\u6838\u5fc3\u4f18\u52bf\u5728\u4e8e\u5176\u6210\u529f\u6574\u5408\u4e86\u89c6\u89c9\u6587\u672c\u7684\u7406\u89e3\u548c\u751f\u6210\u80fd\u529b\u3002\u4f20\u7edf\u7814\u7a76\u4e2d\uff0c\u8fd9\u4e24\u7c7b\u4efb\u52a1\u901a\u5e38\u7531\u72ec\u7acb\u6a21\u578b\u5904\u7406\u3002TextHarmony \u901a\u8fc7\u878d\u5408\u8fd9\u4e24\u5927\u7c7b<mark data-id=\"d92a1cda-8e30-4199-9678-2d8c6c17d736\" data-type=\"tech_methods\">\u751f\u6210\u6a21\u578b<\/mark>\uff0c\u5b9e\u73b0\u4e86\u89c6\u89c9\u6587\u5b57\u7406\u89e3\u548c\u751f\u6210\u7684\u540c\u6b65\u8fdb\u884c\uff0c\u4ece\u800c\u7edf\u7b79\u4e86 OCR \u9886\u57df\u7684\u591a\u6570\u4efb\u52a1\u3002<\/p>\n<p>\u7814\u7a76\u8868\u660e\uff0c\u89c6\u89c9\u7406\u89e3\u548c\u751f\u6210\u4e4b\u95f4\u5b58\u5728\u663e\u8457\u5dee\u5f02\uff0c\u76f4\u63a5\u6574\u5408\u53ef\u80fd\u5bfc\u81f4\u4e25\u91cd\u7684\u6a21\u6001\u4e0d\u4e00\u81f4\u95ee\u9898\u3002\u5177\u4f53\u800c\u8a00\uff0c\u591a\u6a21\u6001<mark data-id=\"d92a1cda-8e30-4199-9678-2d8c6c17d736\" data-type=\"tech_methods\">\u751f\u6210\u6a21\u578b<\/mark>\u5728\u6587\u672c\u751f\u6210\uff08\u89c6\u89c9\u611f\u77e5\u3001\u7406\u89e3\uff09\u548c<mark data-id=\"ba1ea8a1-a705-4f89-8348-ec71bae840dc\" data-type=\"tech_tasks\">\u56fe\u50cf\u751f\u6210<\/mark>\u65b9\u9762\uff0c\u76f8\u8f83\u4e8e\u4e13\u95e8\u7684\u5355\u6a21\u6001\u6a21\u578b\uff0c\u6027\u80fd\u51fa\u73b0\u660e\u663e\u9000\u5316\u3002<\/p>\n<p><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-ac8b9d5661af89c5585e05df775f2de2.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-ac8b9d5661af89c5585e05df775f2de2.png\"><\/a><\/p>\n<p>\u6570\u636e\u663e\u793a\uff0c\u591a\u6a21\u6001<mark data-id=\"d92a1cda-8e30-4199-9678-2d8c6c17d736\" data-type=\"tech_methods\">\u751f\u6210\u6a21\u578b<\/mark>\u5728\u6587\u672c\u751f\u6210\u4efb\u52a1\u4e0a\u8f83\u5355\u6a21\u6001\u6a21\u578b\u6548\u679c\u964d\u4f4e 5%\uff0c<mark data-id=\"ba1ea8a1-a705-4f89-8348-ec71bae840dc\" data-type=\"tech_tasks\">\u56fe\u50cf\u751f\u6210<\/mark>\u4efb\u52a1\u4e0a\u6700\u9ad8\u964d\u4f4e 8%\u3002\u800c TextHarmony \u6210\u529f\u7f13\u89e3\u4e86\u8fd9\u4e00\u95ee\u9898\uff0c\u5176\u5728\u4e24\u7c7b\u4efb\u52a1\u4e0a\u7684\u8868\u73b0\u5747\u63a5\u8fd1\u5355\u6a21\u6001\u4e13\u5bb6\u6a21\u578b\u6c34\u5e73\u3002<\/p>\n<p><strong>\u6280\u672f<\/strong><strong>\u521b\u65b0<\/strong><\/p>\n<p>TextHarmony \u91c7\u7528\u4e86 ViT\u3001MLLM \u548c Diffusion Model \u7684\u7ec4\u5408\u67b6\u6784\uff1a<\/p>\n<p>1.ViT \u8d1f\u8d23\u56fe\u50cf\u5230\u89c6\u89c9 token \u5e8f\u5217\u7684\u8f6c\u6362\u3002<\/p>\n<p>2.MLLM \u5904\u7406\u89c6\u89c9 token \u548c\u6587\u672c token \u7684\u4ea4\u53c9\u5e8f\u5217\uff0c\u8f93\u51fa\u4e24\u7c7b token\uff1a<\/p>\n<p>\u00b7\u6587\u672c token \u7ecf\u6587\u672c\u89e3\u7801\u5668\u8f6c\u5316\u4e3a\u6587\u672c\u8f93\u51fa\u3002<\/p>\n<p>\u00b7\u89c6\u89c9 token \u4e0e\u6587\u672c token \u7ed3\u5408\uff0c\u4f5c\u4e3a Diffusion Model \u7684\u6761\u4ef6\u6307\u5f15\uff0c\u751f\u6210\u76ee\u6807\u56fe\u50cf\u3002<\/p>\n<p>\u8fd9\u79cd\u7ed3\u6784\u5b9e\u73b0\u4e86\u591a\u6a21\u6001\u5185\u5bb9\u7684\u5168\u9762\u7406\u89e3\u4e0e\u751f\u6210\u3002<\/p>\n<p><strong>Slide-LoRA\uff1a\u89e3\u51b3\u65b9\u6848<\/strong><\/p>\n<p>\u4e3a\u514b\u670d\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u7684\u6a21\u6001\u4e0d\u4e00\u81f4\u95ee\u9898\uff0c\u7814\u7a76\u8005\u63d0\u51fa\u4e86 Slide-LoRA \u6280\u672f\u3002\u8be5\u65b9\u6cd5\u901a\u8fc7\u52a8\u6001\u6574\u5408\u6a21\u6001\u7279\u5b9a\u548c\u6a21\u6001\u65e0\u5173\u7684 LoRA\uff08Low-Rank Adaptation\uff09\u4e13\u5bb6\uff0c\u5728\u5355\u4e00\u6a21\u578b\u4e2d\u5b9e\u73b0\u4e86\u56fe\u50cf\u548c\u6587\u672c\u751f\u6210\u7a7a\u95f4\u7684\u90e8\u5206\u89e3\u8026\u3002<\/p>\n<p>Slide-LoRA \u5305\u542b\u4e00\u4e2a\u52a8\u6001\u95e8\u63a7\u7f51\u7edc\u548c\u4e09\u4e2a\u4f4e\u79e9\u5206\u89e3\u6a21\u5757\uff1a<\/p>\n<p>\u00b7\u6a21\u6001\u7279\u5b9a LoRA \u4e13\u5bb6\u805a\u7126\u4e8e\u7279\u5b9a\u6a21\u6001\uff08\u89c6\u89c9\u6216\u8bed\u8a00\uff09\u7684\u751f\u6210\u4efb\u52a1\u3002<\/p>\n<p>\u00b7\u6a21\u6001\u65e0\u5173 LoRA \u4e13\u5bb6\u5904\u7406\u8de8\u6a21\u6001\u7684\u901a\u7528\u7279\u5f81\u3002<\/p>\n<p>\u00b7\u52a8\u6001\u95e8\u63a7\u7f51\u7edc\u6839\u636e\u8f93\u5165\u7279\u5f81\uff0c\u7075\u6d3b\u8c03\u5ea6\u4e0d\u540c\u4e13\u5bb6\u7684\u53c2\u4e0e\u5ea6\u3002<\/p>\n<p><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-aaaaa2c22ac9b634794c5226bbdb3bcf.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-aaaaa2c22ac9b634794c5226bbdb3bcf.png\"><\/a><\/p>\n<h3><strong>DetailedTextCaps-100K: \u9ad8\u8d28\u91cf\u6570\u636e\u96c6<\/strong><\/h3>\n<p>\u4e3a\u63d0\u5347\u89c6\u89c9\u6587\u672c\u751f\u6210\u6027\u80fd\uff0c\u7814\u7a76\u56e2\u961f\u5f00\u53d1\u4e86 DetailedTextCaps-100K \u6570\u636e\u96c6\u3002\u8be5\u96c6\u5229\u7528\u95ed\u6e90 MLLM\uff08Gemini Pro\uff09\u751f\u6210\u8be6\u5c3d\u7684\u56fe\u50cf\u63cf\u8ff0\uff0c\u4e3a\u6a21\u578b\u63d0\u4f9b\u4e86\u66f4\u4e30\u5bcc\u3001\u66f4\u805a\u7126\u4e8e\u89c6\u89c9\u548c\u6587\u672c\u5143\u7d20\u7684\u8bad\u7ec3\u8d44\u6e90\u3002<\/p>\n<p><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-f9ebc9320e1ecd2b1b778332aa4de9da.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-f9ebc9320e1ecd2b1b778332aa4de9da.png\"><\/a><\/p>\n<p><strong>\u8bad\u7ec3\u7b56\u7565<\/strong><\/p>\n<p>TextHarmony \u91c7\u7528\u4e24\u9636\u6bb5\u8bad\u7ec3\u65b9\u6cd5\uff1a<\/p>\n<p>1.\u9996\u9636\u6bb5\u5229\u7528 MARIO-LAION \u548c DocStruct4M \u7b49\u56fe\u6587\u5bf9\u9884\u8bad\u7ec3\u5bf9\u9f50\u6a21\u5757\u548c\u56fe\u50cf\u89e3\u7801\u5668\uff0c\u6784\u5efa\u57fa\u7840\u7684\u6587\u672c\u751f\u6210\u4e0e<mark data-id=\"ba1ea8a1-a705-4f89-8348-ec71bae840dc\" data-type=\"tech_tasks\">\u56fe\u50cf\u751f\u6210<\/mark>\u80fd\u529b\u3002<\/p>\n<p>2.\u6b21\u9636\u6bb5\u8fd0\u7528\u89c6\u89c9\u6587\u672c\u7684\u751f\u6210\u3001\u7f16\u8f91\u3001\u7406\u89e3\u3001\u611f\u77e5\u56db\u7c7b\u6570\u636e\u8fdb\u884c\u7edf\u4e00\u5fae\u8c03\u3002\u6b64\u9636\u6bb5\u5f00\u653e ViT\u3001\u5bf9\u9f50\u6a21\u5757\u3001\u56fe\u50cf\u89e3\u7801\u5668\u548c Slide-LoRA \u7684\u53c2\u6570\u66f4\u65b0\uff0c\u4ee5\u83b7\u5f97\u7edf\u4e00\u7684\u591a\u6a21\u6001\u7406\u89e3\u4e0e\u751f\u6210\u80fd\u529b\u3002<\/p>\n<p><strong>\u5b9e\u9a8c\u8bc4\u4f30<\/strong><\/p>\n<p>\u7814\u7a76\u8005\u5bf9 TextHarmony \u5728\u89c6\u89c9\u6587\u672c\u573a\u666f\u4e0b\u8fdb\u884c\u4e86\u5168\u9762\u8bc4\u4f30\uff0c\u6db5\u76d6\u7406\u89e3\u3001\u611f\u77e5\u3001\u751f\u6210\u4e0e\u7f16\u8f91\u56db\u4e2a\u7ef4\u5ea6\uff1a<\/p>\n<p><\/p>\n<p>1.\u89c6\u89c9\u6587\u672c\u7406\u89e3\uff1aTextHarmony \u663e\u8457\u4f18\u4e8e\u591a\u6a21\u6001<mark data-id=\"d92a1cda-8e30-4199-9678-2d8c6c17d736\" data-type=\"tech_methods\">\u751f\u6210\u6a21\u578b<\/mark>\uff0c\u6027\u80fd\u63a5\u8fd1 Monkey \u7b49\u4e13\u4e1a\u6587\u5b57\u7406\u89e3\u6a21\u578b\u3002<\/p>\n<p><\/p>\n<p><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-e1a70c67074749cc911cf1bcc7abef09.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-e1a70c67074749cc911cf1bcc7abef09.png\"><\/a><\/p>\n<p>2.\u89c6\u89c9\u6587\u672c\u611f\u77e5\uff1a\u5728OCR\u5b9a\u4f4d\u4efb\u52a1\u4e0a\uff0cTextHarmony\u8d85\u8fc7\u4e86TGDoc\u3001DocOwl1.5\u7b49\u77e5\u540d\u6a21\u578b\u3002<\/p>\n<p><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-d5cc2e460fa5fb5bfd691d1aedfc96f9.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-d5cc2e460fa5fb5bfd691d1aedfc96f9.png\"><\/a><\/p>\n<p>3.\u89c6\u89c9\u6587\u672c\u7f16\u8f91\u4e0e\u751f\u6210\uff1aTextHarmony \u5927\u5e45\u9886\u5148\u4e8e\u73b0\u6709\u591a\u6a21\u6001<mark data-id=\"d92a1cda-8e30-4199-9678-2d8c6c17d736\" data-type=\"tech_methods\">\u751f\u6210\u6a21\u578b<\/mark>\uff0c\u4e14\u4e0e TextDiffuser2 \u7b49\u4e13\u4e1a\u6a21\u578b\u76f8\u5f53\u3002<\/p>\n<p>\u3002<\/p>\n<p><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-51e8d0a20c01cc48d8712be33db40ad5.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-51e8d0a20c01cc48d8712be33db40ad5.png\"><\/a><\/p>\n<p><strong>\u6587\u5b57\u751f\u6210\u6548\u679c\u5bf9\u6bd4<\/strong><\/p>\n<p><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-0d9ffef121930676ab36f90fe47fe5c3.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-0d9ffef121930676ab36f90fe47fe5c3.png\"><\/a><\/p>\n<p><strong>\u6587\u5b57\u7f16\u8f91\u6548\u679c\u5bf9\u6bd4<\/strong><\/p>\n<p><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-dfaede9555cad643b07597a8c15019d7.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-dfaede9555cad643b07597a8c15019d7.png\"><\/a><\/p>\n<p><strong>\u6587\u5b57\u56fe\u50cf\u611f\u77e5\u4e0e\u7406\u89e3\u53ef\u89c6\u5316<\/strong><\/p>\n<p><a href=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-506ba610cea3932b4fc7102ba1f6a953.png\" data-fancybox=\"images\" data-fancybox=\"gallery\"><img decoding=\"async\" src=\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-506ba610cea3932b4fc7102ba1f6a953.png\"><\/a><\/p>\n<p><strong>\u603b\u7ed3<\/strong><strong>\u4e0e\u5c55\u671b<\/strong><\/p>\n<p>TextHarmony \u4f5c\u4e3a OCR \u9886\u57df\u7684\u591a\u529f\u80fd\u591a\u6a21\u6001<mark data-id=\"d92a1cda-8e30-4199-9678-2d8c6c17d736\" data-type=\"tech_methods\">\u751f\u6210\u6a21\u578b<\/mark>\uff0c\u6210\u529f\u7edf\u4e00\u4e86\u89c6\u89c9\u6587\u672c\u7406\u89e3\u548c\u751f\u6210\u4efb\u52a1\u3002\u901a\u8fc7\u521b\u65b0\u7684 Slide-LoRA \u6280\u672f\uff0c\u5b83\u6709\u6548\u89e3\u51b3\u4e86\u591a\u6a21\u6001\u751f\u6210\u4e2d\u7684\u6a21\u6001\u4e0d\u4e00\u81f4\u95ee\u9898\uff0c\u5728\u5355\u4e00\u6a21\u578b\u4e2d\u5b9e\u73b0\u4e86\u89c6\u89c9\u4e0e\u8bed\u8a00\u6a21\u6001\u7684\u548c\u8c10\u7edf\u4e00\u3002TextHarmony \u5728\u89c6\u89c9\u6587\u5b57\u611f\u77e5\u3001\u7406\u89e3\u3001\u751f\u6210\u548c\u7f16\u8f91\u65b9\u9762\u5c55\u73b0\u51fa\u5353\u8d8a\u6027\u80fd\uff0c\u4e3a\u590d\u6742\u7684\u89c6\u89c9\u6587\u672c\u4ea4\u4e92\u4efb\u52a1\u5f00\u8f9f\u4e86\u65b0\u7684\u53ef\u80fd\u6027\u3002<\/p>\n<p>\u8fd9\u9879\u7814\u7a76\u4e0d\u4ec5\u63a8\u52a8\u4e86 OCR \u6280\u672f\u7684\u8fdb\u6b65\uff0c\u4e5f\u4e3a<mark data-id=\"2d28aa9c-942d-471d-bd96-8bfefb7144e0\" data-type=\"concepts\">\u4eba\u5de5\u667a\u80fd<\/mark>\u5728\u7406\u89e3\u548c\u521b\u9020\u65b9\u9762\u7684\u53d1\u5c55\u63d0\u4f9b\u4e86\u91cd\u8981\u53c2\u8003\u3002\u672a\u6765\uff0cTextHarmony \u6709\u671b\u5728\u81ea\u52a8\u6587\u6863\u5904\u7406\u3001\u667a\u80fd\u5185\u5bb9\u521b\u4f5c\u3001\u6559\u80b2\u8f85\u52a9\u7b49\u591a\u4e2a\u9886\u57df\u53d1\u6325\u91cd\u8981\u4f5c\u7528\uff0c\u8fdb\u4e00\u6b65\u63a8\u52a8<mark data-id=\"2d28aa9c-942d-471d-bd96-8bfefb7144e0\" data-type=\"concepts\">\u4eba\u5de5\u667a\u80fd<\/mark>\u7684\u5e94\u7528\u3002<\/p>\n<p>\u6587\u7ae0\u6765\u6e90\u4e8e\u4e92\u8054\u7f51:<a href=\"https:\/\/www.jiqizhixin.com\/articles\/2024-10-16\" target=\"_blank\">NeurIPS2024 | OCR-Omni\u6765\u4e86\uff0c\u5b57\u8282&amp;\u534e\u5e08\u63d0\u51fa\u7edf\u4e00\u7684\u591a\u6a21\u6001\u6587\u5b57\u7406\u89e3\u4e0e\u751f\u6210\u5927\u6a21\u578b<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u6587\u7ae0\u6765\u6e90\u4e8e\u4e92\u8054\u7f51:NeurIPS2024 [&hellip;]<\/p>\n","protected":false},"author":3,"featured_media":0,"comment_status":"open","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"categories":[27],"tags":[73,68],"class_list":["post-36127","post","type-post","status-publish","format-standard","hentry","category-news","tag-73","tag-68"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v26.4 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>NeurIPS2024 | OCR-Omni\u6765\u4e86\uff0c\u5b57\u8282&amp;\u534e\u5e08\u63d0\u51fa\u7edf\u4e00\u7684\u591a\u6a21\u6001\u6587\u5b57\u7406\u89e3\u4e0e\u751f\u6210\u5927\u6a21\u578b - \u4e00\u8d77AI\u6280\u672f<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/17aitech.com\/?p=36127\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/17aitech.com\/?p=36127\",\"url\":\"https:\/\/17aitech.com\/?p=36127\",\"name\":\"NeurIPS2024 | OCR-Omni\u6765\u4e86\uff0c\u5b57\u8282&amp;\u534e\u5e08\u63d0\u51fa\u7edf\u4e00\u7684\u591a\u6a21\u6001\u6587\u5b57\u7406\u89e3\u4e0e\u751f\u6210\u5927\u6a21\u578b - \u4e00\u8d77AI\u6280\u672f\",\"isPartOf\":{\"@id\":\"https:\/\/17aitech.com\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/17aitech.com\/?p=36127#primaryimage\"},\"image\":{\"@id\":\"https:\/\/17aitech.com\/?p=36127#primaryimage\"},\"thumbnailUrl\":\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-d80cfc719b4c053dacd27bffbb8f7b8c.png\",\"datePublished\":\"2025-01-06T05:00:11+00:00\",\"author\":{\"@id\":\"https:\/\/17aitech.com\/#\/schema\/person\/60225458499e817ae0af73e67e440b9d\"},\"breadcrumb\":{\"@id\":\"https:\/\/17aitech.com\/?p=36127#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/17aitech.com\/?p=36127\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/17aitech.com\/?p=36127#primaryimage\",\"url\":\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-d80cfc719b4c053dacd27bffbb8f7b8c.png\",\"contentUrl\":\"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-d80cfc719b4c053dacd27bffbb8f7b8c.png\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/17aitech.com\/?p=36127#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/17aitech.com\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"NeurIPS2024 | OCR-Omni\u6765\u4e86\uff0c\u5b57\u8282&amp;\u534e\u5e08\u63d0\u51fa\u7edf\u4e00\u7684\u591a\u6a21\u6001\u6587\u5b57\u7406\u89e3\u4e0e\u751f\u6210\u5927\u6a21\u578b\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/17aitech.com\/#website\",\"url\":\"https:\/\/17aitech.com\/\",\"name\":\"\u4e00\u8d77AI\u6280\u672f\",\"description\":\"\u8ba9AI\u77e5\u8bc6\u89e6\u624b\u53ef\u53ca\",\"alternateName\":\"\u4e00\u8d77AI\u6280\u672f\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/17aitech.com\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/17aitech.com\/#\/schema\/person\/60225458499e817ae0af73e67e440b9d\",\"name\":\"AI\u5c0f\u52a9\u624b\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/17aitech.com\/#\/schema\/person\/image\/\",\"url\":\"\/\/17aitech.com\/wp-content\/uploads\/2024\/04\/robot_3.png\",\"contentUrl\":\"\/\/17aitech.com\/wp-content\/uploads\/2024\/04\/robot_3.png\",\"caption\":\"AI\u5c0f\u52a9\u624b\"},\"description\":\"\u8fd9\u4e2a\u4eba\u5f88\u61d2\uff0c\u4ec0\u4e48\u90fd\u6ca1\u6709\u7559\u4e0b\uff5e\",\"url\":\"https:\/\/17aitech.com\/?page_id=33738&user=3\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"NeurIPS2024 | OCR-Omni\u6765\u4e86\uff0c\u5b57\u8282&amp;\u534e\u5e08\u63d0\u51fa\u7edf\u4e00\u7684\u591a\u6a21\u6001\u6587\u5b57\u7406\u89e3\u4e0e\u751f\u6210\u5927\u6a21\u578b - \u4e00\u8d77AI\u6280\u672f","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/17aitech.com\/?p=36127","schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/17aitech.com\/?p=36127","url":"https:\/\/17aitech.com\/?p=36127","name":"NeurIPS2024 | OCR-Omni\u6765\u4e86\uff0c\u5b57\u8282&amp;\u534e\u5e08\u63d0\u51fa\u7edf\u4e00\u7684\u591a\u6a21\u6001\u6587\u5b57\u7406\u89e3\u4e0e\u751f\u6210\u5927\u6a21\u578b - \u4e00\u8d77AI\u6280\u672f","isPartOf":{"@id":"https:\/\/17aitech.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/17aitech.com\/?p=36127#primaryimage"},"image":{"@id":"https:\/\/17aitech.com\/?p=36127#primaryimage"},"thumbnailUrl":"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-d80cfc719b4c053dacd27bffbb8f7b8c.png","datePublished":"2025-01-06T05:00:11+00:00","author":{"@id":"https:\/\/17aitech.com\/#\/schema\/person\/60225458499e817ae0af73e67e440b9d"},"breadcrumb":{"@id":"https:\/\/17aitech.com\/?p=36127#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/17aitech.com\/?p=36127"]}]},{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/17aitech.com\/?p=36127#primaryimage","url":"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-d80cfc719b4c053dacd27bffbb8f7b8c.png","contentUrl":"https:\/\/17aitech.com\/wp-content\/uploads\/2024\/10\/frc-d80cfc719b4c053dacd27bffbb8f7b8c.png"},{"@type":"BreadcrumbList","@id":"https:\/\/17aitech.com\/?p=36127#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/17aitech.com\/"},{"@type":"ListItem","position":2,"name":"NeurIPS2024 | OCR-Omni\u6765\u4e86\uff0c\u5b57\u8282&amp;\u534e\u5e08\u63d0\u51fa\u7edf\u4e00\u7684\u591a\u6a21\u6001\u6587\u5b57\u7406\u89e3\u4e0e\u751f\u6210\u5927\u6a21\u578b"}]},{"@type":"WebSite","@id":"https:\/\/17aitech.com\/#website","url":"https:\/\/17aitech.com\/","name":"\u4e00\u8d77AI\u6280\u672f","description":"\u8ba9AI\u77e5\u8bc6\u89e6\u624b\u53ef\u53ca","alternateName":"\u4e00\u8d77AI\u6280\u672f","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/17aitech.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/17aitech.com\/#\/schema\/person\/60225458499e817ae0af73e67e440b9d","name":"AI\u5c0f\u52a9\u624b","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/17aitech.com\/#\/schema\/person\/image\/","url":"\/\/17aitech.com\/wp-content\/uploads\/2024\/04\/robot_3.png","contentUrl":"\/\/17aitech.com\/wp-content\/uploads\/2024\/04\/robot_3.png","caption":"AI\u5c0f\u52a9\u624b"},"description":"\u8fd9\u4e2a\u4eba\u5f88\u61d2\uff0c\u4ec0\u4e48\u90fd\u6ca1\u6709\u7559\u4e0b\uff5e","url":"https:\/\/17aitech.com\/?page_id=33738&user=3"}]}},"_links":{"self":[{"href":"https:\/\/17aitech.com\/index.php?rest_route=\/wp\/v2\/posts\/36127","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/17aitech.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/17aitech.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/17aitech.com\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/17aitech.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=36127"}],"version-history":[{"count":0,"href":"https:\/\/17aitech.com\/index.php?rest_route=\/wp\/v2\/posts\/36127\/revisions"}],"wp:attachment":[{"href":"https:\/\/17aitech.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=36127"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/17aitech.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=36127"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/17aitech.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=36127"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}