{"version":2,"relationshipURLBase":"https://codingai-lec03.pages.dev/course/notebook-blocks/3dc8766f20144da28bbef37cade18a3d.html","relationshipResolution":"Resolve relative relationship URLs against relationshipURLBase; complete records and absolute URLs are in blocks.","title":"来源与下一条问题","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/3dc8766f20144da28bbef37cade18a3d.html","markdown":"https://codingai-lec03.pages.dev/course/notebook-blocks/3dc8766f20144da28bbef37cade18a3d.md","pages":[{"id":"reading-opening","title":"来源与下一条问题","unit":"来源与下一条问题","unitId":"notebook-blocks/3dc8766f20144da28bbef37cade18a3d","chapter":"course","number":1}],"aliases":[{"oldId":"b-db688ea8614f26ddf59dad11","targetId":"b-db688ea8614f26ddf59dad11","page":"reading-opening","operation":"move","beforeSHA256":"68ab73f0c1cbd872efb5dd565e57f07f1b4a2b56c5ea0ce2c6ed91ff60ff5a3f","afterSHA256":"68ab73f0c1cbd872efb5dd565e57f07f1b4a2b56c5ea0ce2c6ed91ff60ff5a3f","reason":"Same source body in authored semantic page"},{"oldId":"b-eeea00bfbe1a2f4a93478392","targetId":"b-eeea00bfbe1a2f4a93478392","page":"reading-opening","operation":"move","beforeSHA256":"0b1b3ee44f7e3b19e19ef4416d120b089156bc5f82ce8a5f77094b73ee8dd5b4","afterSHA256":"0b1b3ee44f7e3b19e19ef4416d120b089156bc5f82ce8a5f77094b73ee8dd5b4","reason":"Same source body in authored semantic page"},{"oldId":"b-420d3d3452fb28bddac8137c","targetId":"b-420d3d3452fb28bddac8137c","page":"reading-opening","operation":"move","beforeSHA256":"42cba31deccba2a350b9b23bd0ca1dcaa4c4c5cb9d35c773e1fcb89124788ccd","afterSHA256":"42cba31deccba2a350b9b23bd0ca1dcaa4c4c5cb9d35c773e1fcb89124788ccd","reason":"Same source body in authored semantic page"},{"oldId":"b-2256870d3021d461592450b6","targetId":"b-2256870d3021d461592450b6","page":"reading-opening","operation":"move","beforeSHA256":"b1c765d7bf903321abd36e75bae49048d61faec2aa3fd11922b328c5970003b1","afterSHA256":"b1c765d7bf903321abd36e75bae49048d61faec2aa3fd11922b328c5970003b1","reason":"Same source body in authored semantic page"},{"oldId":"b-050b0dc41374a4c14a8bb63b","targetId":"b-050b0dc41374a4c14a8bb63b","page":"reading-opening","operation":"move","beforeSHA256":"26a7cbe49a363a14a248834129e6d61a20d9aa9861079f66efe0ecca0a3f9a6b","afterSHA256":"26a7cbe49a363a14a248834129e6d61a20d9aa9861079f66efe0ecca0a3f9a6b","reason":"Same source body in authored semantic page"},{"oldId":"b-5658dc2628d390b683267801","targetId":"b-5658dc2628d390b683267801","page":"reading-opening","operation":"move","beforeSHA256":"80d65a4fb686786c0f111c5580f2a1dc3c9a542aed8d416cdf9ab7f723b2f3eb","afterSHA256":"80d65a4fb686786c0f111c5580f2a1dc3c9a542aed8d416cdf9ab7f723b2f3eb","reason":"Same source body in authored semantic page"},{"oldId":"b-90be9ddc59d7038cb20ddb26","targetId":"b-90be9ddc59d7038cb20ddb26","page":"reading-opening","operation":"move","beforeSHA256":"54419b0c3585352e9f49dbf54b576784a7f8a9b5d3df629a511b72a9ef3055ec","afterSHA256":"54419b0c3585352e9f49dbf54b576784a7f8a9b5d3df629a511b72a9ef3055ec","reason":"Same source body in authored semantic page"},{"oldId":"block-3dc8766f20144da28bbef37cade18a3d--来源与下一条问题","targetId":"b-050b0dc41374a4c14a8bb63b","operation":"retain-source-identity"}],"blocks":[{"id":"b-db688ea8614f26ddf59dad11","type":"p","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/3dc8766f20144da28bbef37cade18a3d.html#b-db688ea8614f26ddf59dad11","content":"[LAB 10 · 从 RGB 图像训练小型检测器](/course/notebooks/a14237bfa9dd413d959ab4e458e11f42.html)","sha256":"68ab73f0c1cbd872efb5dd565e57f07f1b4a2b56c5ea0ce2c6ed91ff60ff5a3f","chapter":{"id":"course","title":"来源与下一条问题"},"unit":{"id":"notebook-blocks/3dc8766f20144da28bbef37cade18a3d","title":"来源与下一条问题"},"page":{"id":"reading-opening","title":"来源与下一条问题","number":1},"sectionTitle":"来源与下一条问题","currentQuestion":"来源与下一条问题","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":null,"nextBlock":"b-eeea00bfbe1a2f4a93478392","prerequisites":[],"relatedBlocks":[]},{"id":"b-eeea00bfbe1a2f4a93478392","type":"h1","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/3dc8766f20144da28bbef37cade18a3d.html#b-eeea00bfbe1a2f4a93478392","content":"# 来源与下一条问题","sha256":"0b1b3ee44f7e3b19e19ef4416d120b089156bc5f82ce8a5f77094b73ee8dd5b4","chapter":{"id":"course","title":"来源与下一条问题"},"unit":{"id":"notebook-blocks/3dc8766f20144da28bbef37cade18a3d","title":"来源与下一条问题"},"page":{"id":"reading-opening","title":"来源与下一条问题","number":1},"sectionTitle":"来源与下一条问题","currentQuestion":"来源与下一条问题","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-db688ea8614f26ddf59dad11","nextBlock":"b-420d3d3452fb28bddac8137c","prerequisites":[{"id":"b-db688ea8614f26ddf59dad11","url":"#b-db688ea8614f26ddf59dad11","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-420d3d3452fb28bddac8137c","type":"p","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/3dc8766f20144da28bbef37cade18a3d.html#b-420d3d3452fb28bddac8137c","content":"主线阅读与优先实验","sha256":"42cba31deccba2a350b9b23bd0ca1dcaa4c4c5cb9d35c773e1fcb89124788ccd","chapter":{"id":"course","title":"来源与下一条问题"},"unit":{"id":"notebook-blocks/3dc8766f20144da28bbef37cade18a3d","title":"来源与下一条问题"},"page":{"id":"reading-opening","title":"来源与下一条问题","number":1},"sectionTitle":"来源与下一条问题","currentQuestion":"来源与下一条问题","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-eeea00bfbe1a2f4a93478392","nextBlock":"b-2256870d3021d461592450b6","prerequisites":[{"id":"b-eeea00bfbe1a2f4a93478392","url":"#b-eeea00bfbe1a2f4a93478392","role":"section-title"},{"id":"b-db688ea8614f26ddf59dad11","url":"#b-db688ea8614f26ddf59dad11","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-2256870d3021d461592450b6","type":"p","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/3dc8766f20144da28bbef37cade18a3d.html#b-2256870d3021d461592450b6","content":"RGB图像、单特征层锚框检测器、分类/位置损失、NMS、独立检测评价","sha256":"b1c765d7bf903321abd36e75bae49048d61faec2aa3fd11922b328c5970003b1","chapter":{"id":"course","title":"来源与下一条问题"},"unit":{"id":"notebook-blocks/3dc8766f20144da28bbef37cade18a3d","title":"来源与下一条问题"},"page":{"id":"reading-opening","title":"来源与下一条问题","number":1},"sectionTitle":"来源与下一条问题","currentQuestion":"来源与下一条问题","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-420d3d3452fb28bddac8137c","nextBlock":"b-050b0dc41374a4c14a8bb63b","prerequisites":[{"id":"b-eeea00bfbe1a2f4a93478392","url":"#b-eeea00bfbe1a2f4a93478392","role":"section-title"},{"id":"b-db688ea8614f26ddf59dad11","url":"#b-db688ea8614f26ddf59dad11","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-050b0dc41374a4c14a8bb63b","type":"h2","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/3dc8766f20144da28bbef37cade18a3d.html#b-050b0dc41374a4c14a8bb63b","content":"## 来源与下一条问题","sha256":"26a7cbe49a363a14a248834129e6d61a20d9aa9861079f66efe0ecca0a3f9a6b","chapter":{"id":"course","title":"来源与下一条问题"},"unit":{"id":"notebook-blocks/3dc8766f20144da28bbef37cade18a3d","title":"来源与下一条问题"},"page":{"id":"reading-opening","title":"来源与下一条问题","number":1},"sectionTitle":"来源与下一条问题","currentQuestion":"来源与下一条问题","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-2256870d3021d461592450b6","nextBlock":"b-5658dc2628d390b683267801","prerequisites":[{"id":"b-eeea00bfbe1a2f4a93478392","url":"#b-eeea00bfbe1a2f4a93478392","role":"section-title"},{"id":"b-db688ea8614f26ddf59dad11","url":"#b-db688ea8614f26ddf59dad11","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-5658dc2628d390b683267801","type":"p","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/3dc8766f20144da28bbef37cade18a3d.html#b-5658dc2628d390b683267801","content":"2026-09-22 实时核对 D2L 1.0.3：[Anchor Boxes §14.4.3–14.4.4](https://d2l.ai/chapter_computer-vision/anchor.html#labeling-anchor-boxes-in-training-data) 用于锚框分配、偏移与 NMS；[SSD §14.7.1–14.7.3](https://d2l.ai/chapter_computer-vision/ssd.html)用于共享骨干、分类/回归头与完整训练路径。这里独立实现单特征层、小合成场景、ignore 区间、hard-negative CE、Smooth L1 与全数据 AP；没有把 D2L 的香蕉数据或论文成绩说成本次结果。","sha256":"80d65a4fb686786c0f111c5580f2a1dc3c9a542aed8d416cdf9ab7f723b2f3eb","chapter":{"id":"course","title":"来源与下一条问题"},"unit":{"id":"notebook-blocks/3dc8766f20144da28bbef37cade18a3d","title":"来源与下一条问题"},"page":{"id":"reading-opening","title":"来源与下一条问题","number":1},"sectionTitle":"来源与下一条问题","currentQuestion":"来源与下一条问题","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-050b0dc41374a4c14a8bb63b","nextBlock":"b-90be9ddc59d7038cb20ddb26","prerequisites":[{"id":"b-050b0dc41374a4c14a8bb63b","url":"#b-050b0dc41374a4c14a8bb63b","role":"section-title"},{"id":"b-db688ea8614f26ddf59dad11","url":"#b-db688ea8614f26ddf59dad11","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-90be9ddc59d7038cb20ddb26","type":"p","scope":"document","url":"https://codingai-lec03.pages.dev/course/notebook-blocks/3dc8766f20144da28bbef37cade18a3d.html#b-90be9ddc59d7038cb20ddb26","content":"分类给出“是什么”，检测再加“各在何处”。即使框完全正确，矩形框内仍包含椭圆四角的背景；LAB11 接着将每个像素列为分类位置，检查边界、细小对象和采样损失。","sha256":"54419b0c3585352e9f49dbf54b576784a7f8a9b5d3df629a511b72a9ef3055ec","chapter":{"id":"course","title":"来源与下一条问题"},"unit":{"id":"notebook-blocks/3dc8766f20144da28bbef37cade18a3d","title":"来源与下一条问题"},"page":{"id":"reading-opening","title":"来源与下一条问题","number":1},"sectionTitle":"来源与下一条问题","currentQuestion":"来源与下一条问题","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-5658dc2628d390b683267801","nextBlock":null,"prerequisites":[{"id":"b-050b0dc41374a4c14a8bb63b","url":"#b-050b0dc41374a4c14a8bb63b","role":"section-title"},{"id":"b-db688ea8614f26ddf59dad11","url":"#b-db688ea8614f26ddf59dad11","role":"problem-context"}],"relatedBlocks":[]}]}