
{"id":10771,"date":"2026-08-05T18:26:23","date_gmt":"2026-08-05T10:26:23","guid":{"rendered":"https:\/\/infernews.com\/blog\/joyai-video-edit\/"},"modified":"2026-08-05T18:45:37","modified_gmt":"2026-08-05T10:45:37","slug":"joyai-video-edit","status":"publish","type":"post","link":"https:\/\/infernews.com\/blog\/joyai-video-edit\/","title":{"rendered":"JoyAI \u628a\u5373\u6642\u4e32\u6d41\u5f71\u7247\u7de8\u8f2f\u63a8\u5411 720p"},"content":{"rendered":"\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/uploads\/2026\/08\/teaser-1.jpg\" alt=\"JoyAI-Video-Edit teaser\"><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">\u76f4\u64ad\u756b\u9762\u6216\u4e0a\u8f09\u5f71\u7247\u4e0d\u5fc5\u7b49\u5230\u5b8c\u6574\u7247\u6bb5\u6e96\u5099\u597d\uff0c\u4fbf\u80fd\u4e00\u908a\u8f38\u5165\u81ea\u7136\u8a9e\u8a00\u6307\u4ee4\u3001\u4e00\u908a\u770b\u5230\u4fee\u6539\u7d50\u679c\u3002JoyAI-Video-Edit \u5c6c\u65bc\u958b\u6e90\u5f71\u7247\u751f\u6210\u53ca\u5f71\u50cf\u8655\u7406\u6a21\u578b\uff0c\u8655\u7406\u7684\u662f\u5f71\u7247\u4e32\u6d41\u4e2d\u5ef6\u9072\u9ad8\u3001\u5fc5\u9808\u9810\u5148\u77e5\u9053\u7247\u9577\uff0c\u4ee5\u53ca\u96e3\u4ee5\u7dad\u6301\u9023\u8cab\u6027\u7684\u7de8\u8f2f\u6d41\u7a0b\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u5b83\u652f\u63f4\u4e3b\u9ad4\u4fee\u6539\u3001\u5c40\u90e8\u5340\u57df\u8abf\u6574\u3001\u80cc\u666f\u66ff\u63db\u3001\u98a8\u683c\u8f49\u63db\u3001\u52d5\u4f5c\u6539\u8b8a\u548c\u53c3\u8003\u5f71\u50cf\u5f15\u5c0e\uff0c\u9069\u5408\u4e92\u52d5\u793a\u7bc4\u3001\u76f4\u64ad\u6548\u679c\u53ca\u9700\u8981\u5373\u6642\u9810\u89bd\u7684\u5275\u4f5c\u5de5\u5177\u3002\u7cfb\u7d71\u4ee5 Multimodal Large Language Model\uff08MLLM\uff09\u689d\u4ef6\u7de8\u78bc\u5668\u3001causal video Variational Autoencoder\uff08VAE\uff09\u53ca 16B-parameter Multimodal Diffusion Transformer\uff08MMDiT\uff09\u7d44\u6210\uff0c\u9010\u6bb5\u8655\u7406\u65b0\u62b5\u9054\u7684\u756b\u9762\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u540c\u985e\u5f71\u7247\u751f\u6210\u65b9\u6cd5\u5f80\u5f80\u5148\u53d6\u5f97\u5b8c\u6574\u5f71\u7247\uff0c\u518d\u4e00\u6b21\u904e\u9032\u884c\u96e2\u7dda\u8655\u7406\uff1bJoyAI-Video-Edit \u6539\u7528 autoregressive diffusion\uff0c\u914d\u5408 aligned autoregressive distribution matching distillation\u3001long-horizon optimization\u3001bounded Key-Value state\uff08KV-state\uff09inference \u548c deployment-oriented scheduling\uff0c\u63db\u53d6\u4e32\u6d41\u901f\u5ea6\u3002\u4e0d\u904e\uff0c\u9019\u7a2e\u8a2d\u8a08\u4ecd\u8981\u7559\u610f\u9577\u6642\u9593\u8f38\u51fa\u53ef\u80fd\u51fa\u73fe\u7684 temporal drift\uff0c\u800c\u4e14\u6d88\u8cbb\u7d1a GPU \u652f\u63f4\u4ecd\u5217\u70ba\u5f85\u8fa6\u5de5\u4f5c\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u90e8\u7f72\u57fa\u6e96\u5728 720&#215;1280 \u89e3\u50cf\u5ea6\u9054\u5230 30.19 FPS end-to-end throughput\uff0c\u4ee3\u8868\u7cfb\u7d71\u5df2\u63a5\u8fd1\u4e92\u52d5\u5f0f\u5f71\u7247\u8655\u7406\u6240\u9700\u7684\u901f\u5ea6\uff0c\u4f46\u4e0d\u80fd\u76f4\u63a5\u7b49\u540c\u65bc\u6240\u6709\u786c\u4ef6\u548c\u6307\u4ee4\u4e0b\u90fd\u80fd\u4fdd\u6301\u76f8\u540c\u8868\u73fe\u3002Hugging Face \u63d0\u4f9b JoyAI-Video-Edit checkpoint\uff0cGitHub \u540c\u6642\u63d0\u4f9b\u90e8\u7f72\u7a0b\u5f0f\u78bc\u548c\u7dda\u4e0a Demo\uff0c\u8f03\u9069\u5408\u5177\u5099 GPU \u8cc7\u6e90\u3001\u5e0c\u671b\u6574\u5408\u5f71\u7247\u5de5\u4f5c\u6d41\uff0c\u6216\u7814\u7a76 Computer Vision \u8207\u4e32\u6d41\u751f\u6210\u7684\u5718\u968a\u3002<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>\u5373\u6642\u4e32\u6d41\uff1a<\/strong>\u756b\u9762\u9010\u5e40\u8655\u7406\uff0c\u4e0d\u8981\u6c42\u9810\u5148\u63d0\u4f9b\u5b8c\u6574\u5f71\u7247\u6216\u56fa\u5b9a\u7247\u9577\u3002<\/li>\n<li><strong>\u6307\u4ee4\u7bc4\u570d\u5ee3\uff1a<\/strong>\u6db5\u84cb\u4e3b\u9ad4\u3001\u5c40\u90e8\u3001\u80cc\u666f\u3001\u98a8\u683c\u3001\u52d5\u4f5c\u53ca\u53c3\u8003\u5f71\u50cf\u7de8\u8f2f\u3002<\/li>\n<li><strong>\u901f\u5ea6\u6307\u6a19\uff1a<\/strong>720&#215;1280 \u4e0b\u9054\u5230 30.19 FPS \u7684\u5b8c\u6574\u6d41\u7a0b\u541e\u5410\u91cf\u3002<\/li>\n<li><strong>\u90e8\u7f72\u53d6\u6368\uff1a<\/strong>\u900f\u904e bounded KV-state inference \u63a7\u5236\u8a08\u7b97\u91cf\uff0c\u4f46\u6d88\u8cbb\u7d1a GPU \u652f\u63f4\u4ecd\u672a\u5b8c\u6210\u3002<\/li>\n<li><strong>\u9069\u7528\u4eba\u7fa4\uff1a<\/strong>\u5f71\u7247\u5de5\u5177\u958b\u767c\u8005\u3001\u76f4\u64ad\u5275\u4f5c\u8005\u53ca\u9700\u8981\u5373\u6642\u9810\u89bd\u7684\u7814\u7a76\u5718\u968a\u3002<\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/jd-opensource\/JoyAI-Video-Edit\" rel=\"noopener noreferrer\"><strong>GitHub<\/strong><\/a> \u00b7 <a href=\"https:\/\/huggingface.co\/jdopensource\/JoyAI-Video-Edit\" rel=\"noopener noreferrer\"><strong>\u6a21\u578b<\/strong><\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u5f71\u7247\u9010\u5e40\u62b5\u9054\u4fbf\u53ef\u6309\u6587\u5b57\u6307\u4ee4\u4fee\u6539\uff0cJoyAI-Video-Edit \u4ee5 30.19 FPS \u9023\u63a5\u5373\u6642\u651d\u5f71\u8207\u751f\u6210\u5f0f\u7de8\u8f2f\u3002<\/p>\n","protected":false},"author":8,"featured_media":10770,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"ai_generated_summary":"","footnotes":""},"categories":[133,178,179,31,119,149,141,187,199],"tags":[],"class_list":["post-10771","post","type-post","status-publish","format-standard","hentry","category-133","category-google","category-nvidia","category-video","category-119","category-149","category-141","category-187","category-dataset-"],"_links":{"self":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/10771","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/comments?post=10771"}],"version-history":[{"count":1,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/10771\/revisions"}],"predecessor-version":[{"id":10778,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/10771\/revisions\/10778"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media\/10770"}],"wp:attachment":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media?parent=10771"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/categories?post=10771"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/tags?post=10771"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}