{"version":2,"relationshipURLBase":"https://codingai-lec03.pages.dev/course/notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103.html","relationshipResolution":"Resolve relative relationship URLs against relationshipURLBase; complete records and absolute URLs are in blocks.","title":"前一部分用框和点指定对象","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103.html","markdown":"https://codingai-lec03.pages.dev/course/notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103.md","pages":[{"id":"reading-opening","title":"前一部分用框和点指定对象","unit":"前一部分用框和点指定对象","unitId":"notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103","chapter":"course","number":1}],"aliases":[{"oldId":"b-17b3a5cff02ff7e4fc221d35","targetId":"b-17b3a5cff02ff7e4fc221d35","page":"reading-opening","operation":"move","beforeSHA256":"343fb71807f077069b41c6495a8824ea4d353c740175b5775ccc5145b3bf19ca","afterSHA256":"343fb71807f077069b41c6495a8824ea4d353c740175b5775ccc5145b3bf19ca","reason":"Same source body in authored semantic page"},{"oldId":"b-a024b572ce0a7d3d93a21e88","targetId":"b-a024b572ce0a7d3d93a21e88","page":"reading-opening","operation":"move","beforeSHA256":"931bb79b97e8854b2d51fad4f0ffb553b3d682273edaf57ae594b01b28965f1e","afterSHA256":"931bb79b97e8854b2d51fad4f0ffb553b3d682273edaf57ae594b01b28965f1e","reason":"Same source body in authored semantic page"},{"oldId":"b-70942c177e49e8362ece7a5b","targetId":"b-70942c177e49e8362ece7a5b","page":"reading-opening","operation":"move","beforeSHA256":"42cba31deccba2a350b9b23bd0ca1dcaa4c4c5cb9d35c773e1fcb89124788ccd","afterSHA256":"42cba31deccba2a350b9b23bd0ca1dcaa4c4c5cb9d35c773e1fcb89124788ccd","reason":"Same source body in authored semantic page"},{"oldId":"b-563a2de31bd468ff8e7cb35f","targetId":"b-563a2de31bd468ff8e7cb35f","page":"reading-opening","operation":"move","beforeSHA256":"41c70c2c07b82678ff0a60f2d009b6afa750c5c691d109e754a69abf23868b50","afterSHA256":"41c70c2c07b82678ff0a60f2d009b6afa750c5c691d109e754a69abf23868b50","reason":"Same source body in authored semantic page"},{"oldId":"b-450aa1829c2ce35075d9a731","targetId":"b-450aa1829c2ce35075d9a731","page":"reading-opening","operation":"move","beforeSHA256":"6ad34e4ada835ae6143061c572cc8edb19103f364741d4cf627501c9ba2aa759","afterSHA256":"6ad34e4ada835ae6143061c572cc8edb19103f364741d4cf627501c9ba2aa759","reason":"Same source body in authored semantic page"},{"oldId":"b-06ffee6d8db4814ac39a66da","targetId":"b-06ffee6d8db4814ac39a66da","page":"reading-opening","operation":"rewrite","beforeSHA256":"fc65d4e0d64271d306a1dcd0daafa1a8009c0e1ba58b2e34d56608e5e7f18cc4","afterSHA256":"7568784e0cb35e602743541c7257dcc26131f360b079886e0148b7b750e1d9b8","reason":"Same source identity with figure semantics or reading context enriched"},{"oldId":"b-0e49bd3b1f3c9700d54460cd","targetId":"b-0e49bd3b1f3c9700d54460cd","page":"reading-opening","operation":"move","beforeSHA256":"c5a151ba290050d898b8b7b1ee06d5e932083a9a348688238f6e032b81d679f2","afterSHA256":"c5a151ba290050d898b8b7b1ee06d5e932083a9a348688238f6e032b81d679f2","reason":"Same source body in authored semantic page"}],"blocks":[{"id":"b-17b3a5cff02ff7e4fc221d35","type":"p","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103.html#b-17b3a5cff02ff7e4fc221d35","content":"[CLS 03 · 视觉与语言：任务、训练与接口](/course/notebooks/680da245145447dd88983aabd136fd24.html)","sha256":"343fb71807f077069b41c6495a8824ea4d353c740175b5775ccc5145b3bf19ca","chapter":{"id":"course","title":"前一部分用框和点指定对象"},"unit":{"id":"notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103","title":"前一部分用框和点指定对象"},"page":{"id":"reading-opening","title":"前一部分用框和点指定对象","number":1},"sectionTitle":"前一部分用框和点指定对象","currentQuestion":"前一部分用框和点指定对象","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":null,"nextBlock":"b-a024b572ce0a7d3d93a21e88","prerequisites":[],"relatedBlocks":[]},{"id":"b-a024b572ce0a7d3d93a21e88","type":"h1","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103.html#b-a024b572ce0a7d3d93a21e88","content":"# 前一部分用框和点指定对象","sha256":"931bb79b97e8854b2d51fad4f0ffb553b3d682273edaf57ae594b01b28965f1e","chapter":{"id":"course","title":"前一部分用框和点指定对象"},"unit":{"id":"notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103","title":"前一部分用框和点指定对象"},"page":{"id":"reading-opening","title":"前一部分用框和点指定对象","number":1},"sectionTitle":"前一部分用框和点指定对象","currentQuestion":"前一部分用框和点指定对象","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-17b3a5cff02ff7e4fc221d35","nextBlock":"b-70942c177e49e8362ece7a5b","prerequisites":[{"id":"b-17b3a5cff02ff7e4fc221d35","url":"#b-17b3a5cff02ff7e4fc221d35","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-70942c177e49e8362ece7a5b","type":"p","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103.html#b-70942c177e49e8362ece7a5b","content":"主线阅读与优先实验","sha256":"42cba31deccba2a350b9b23bd0ca1dcaa4c4c5cb9d35c773e1fcb89124788ccd","chapter":{"id":"course","title":"前一部分用框和点指定对象"},"unit":{"id":"notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103","title":"前一部分用框和点指定对象"},"page":{"id":"reading-opening","title":"前一部分用框和点指定对象","number":1},"sectionTitle":"前一部分用框和点指定对象","currentQuestion":"前一部分用框和点指定对象","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-a024b572ce0a7d3d93a21e88","nextBlock":"b-563a2de31bd468ff8e7cb35f","prerequisites":[{"id":"b-a024b572ce0a7d3d93a21e88","url":"#b-a024b572ce0a7d3d93a21e88","role":"section-title"},{"id":"b-17b3a5cff02ff7e4fc221d35","url":"#b-17b3a5cff02ff7e4fc221d35","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-563a2de31bd468ff8e7cb35f","type":"p","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103.html#b-563a2de31bd468ff8e7cb35f","content":"caption历史、VQA、CLIP候选、关系控制、GLIP/LLaVA与生成桥梁","sha256":"41c70c2c07b82678ff0a60f2d009b6afa750c5c691d109e754a69abf23868b50","chapter":{"id":"course","title":"前一部分用框和点指定对象"},"unit":{"id":"notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103","title":"前一部分用框和点指定对象"},"page":{"id":"reading-opening","title":"前一部分用框和点指定对象","number":1},"sectionTitle":"前一部分用框和点指定对象","currentQuestion":"前一部分用框和点指定对象","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-70942c177e49e8362ece7a5b","nextBlock":"b-450aa1829c2ce35075d9a731","prerequisites":[{"id":"b-a024b572ce0a7d3d93a21e88","url":"#b-a024b572ce0a7d3d93a21e88","role":"section-title"},{"id":"b-17b3a5cff02ff7e4fc221d35","url":"#b-17b3a5cff02ff7e4fc221d35","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-450aa1829c2ce35075d9a731","type":"p","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103.html#b-450aa1829c2ce35075d9a731","content":"前一部分用框和点指定对象。现在把任务说出来：“描述这张桌面图”“找到红杯”“谁拿着球”。这些句子对系统提出的输出要求不同。","sha256":"6ad34e4ada835ae6143061c572cc8edb19103f364741d4cf627501c9ba2aa759","chapter":{"id":"course","title":"前一部分用框和点指定对象"},"unit":{"id":"notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103","title":"前一部分用框和点指定对象"},"page":{"id":"reading-opening","title":"前一部分用框和点指定对象","number":1},"sectionTitle":"前一部分用框和点指定对象","currentQuestion":"前一部分用框和点指定对象","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-563a2de31bd468ff8e7cb35f","nextBlock":"b-06ffee6d8db4814ac39a66da","prerequisites":[{"id":"b-a024b572ce0a7d3d93a21e88","url":"#b-a024b572ce0a7d3d93a21e88","role":"section-title"},{"id":"b-17b3a5cff02ff7e4fc221d35","url":"#b-17b3a5cff02ff7e4fc221d35","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-06ffee6d8db4814ac39a66da","type":"table","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103.html#b-06ffee6d8db4814ac39a66da","content":"| 语言的作用 | 本部分追问 | 需要建立的接口 |\n| --- | --- | --- |\n| 作为输出 | 如何利用图像和已生成词，描述所见？ | 视觉表示连接词序列预测 |\n| 作为学习信号 | 图文配对怎样教会匹配与检索？ | 图像与文本进入可比较空间 |\n| 作为任务条件 | 这句话指哪个区域，或要求回答什么？ | 区域—词语对应，或视觉条件下的指令作答 |","sha256":"7568784e0cb35e602743541c7257dcc26131f360b079886e0148b7b750e1d9b8","chapter":{"id":"course","title":"前一部分用框和点指定对象"},"unit":{"id":"notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103","title":"前一部分用框和点指定对象"},"page":{"id":"reading-opening","title":"前一部分用框和点指定对象","number":1},"sectionTitle":"前一部分用框和点指定对象","currentQuestion":"前一部分用框和点指定对象","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"table":{"headers":["语言的作用","本部分追问","需要建立的接口"],"rows":[["语言的作用","本部分追问","需要建立的接口"],["作为输出","如何利用图像和已生成词，描述所见？","视觉表示连接词序列预测"],["作为学习信号","图文配对怎样教会匹配与检索？","图像与文本进入可比较空间"],["作为任务条件","这句话指哪个区域，或要求回答什么？","区域—词语对应，或视觉条件下的指令作答"]]},"previousBlock":"b-450aa1829c2ce35075d9a731","nextBlock":"b-0e49bd3b1f3c9700d54460cd","prerequisites":[{"id":"b-a024b572ce0a7d3d93a21e88","url":"#b-a024b572ce0a7d3d93a21e88","role":"section-title"},{"id":"b-17b3a5cff02ff7e4fc221d35","url":"#b-17b3a5cff02ff7e4fc221d35","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-0e49bd3b1f3c9700d54460cd","type":"p","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103.html#b-0e49bd3b1f3c9700d54460cd","content":"这不是互斥的三类系统，也不是后一阶段淘汰前一阶段。一个视觉助手可以使用图文配对学习，用问题指定任务，再生成文字回答。先明确语言此刻的作用，再讨论需要的训练与接口。","sha256":"c5a151ba290050d898b8b7b1ee06d5e932083a9a348688238f6e032b81d679f2","chapter":{"id":"course","title":"前一部分用框和点指定对象"},"unit":{"id":"notebook-blocks/7ae8158cd82e44c29bd3cc05ccee7103","title":"前一部分用框和点指定对象"},"page":{"id":"reading-opening","title":"前一部分用框和点指定对象","number":1},"sectionTitle":"前一部分用框和点指定对象","currentQuestion":"前一部分用框和点指定对象","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-06ffee6d8db4814ac39a66da","nextBlock":null,"prerequisites":[{"id":"b-a024b572ce0a7d3d93a21e88","url":"#b-a024b572ce0a7d3d93a21e88","role":"section-title"},{"id":"b-17b3a5cff02ff7e4fc221d35","url":"#b-17b3a5cff02ff7e4fc221d35","role":"problem-context"}],"relatedBlocks":[]}]}
