{"version":2,"relationshipURLBase":"https://codingai-lec03.pages.dev/course/blocks/mask-iou-worked-example.html","relationshipResolution":"Resolve relative relationship URLs against relationshipURLBase; complete records and absolute URLs are in blocks.","title":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","url":"https://codingai-lec03.pages.dev/course/blocks/mask-iou-worked-example.html","markdown":"https://codingai-lec03.pages.dev/course/blocks/mask-iou-worked-example.md","pages":[{"id":"reading-opening","title":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","unit":"怎样直接学习一张像素类别地图？","unitId":"pixel-prediction-objective","chapter":"II","number":1}],"aliases":[{"oldId":"b-84341872df1d64d15894f412","targetId":"b-84341872df1d64d15894f412","page":"reading-opening","operation":"move","beforeSHA256":"7c9539c84882d62891054298b4186066b7eed88e53712edb4ab393fee68e5242","afterSHA256":"7c9539c84882d62891054298b4186066b7eed88e53712edb4ab393fee68e5242","reason":"Same source body in authored semantic page"},{"oldId":"b-d009c369797d872bc29b9268","targetId":"b-d009c369797d872bc29b9268","page":"reading-opening","operation":"move","beforeSHA256":"1c12d6c2d1f8cd98252583075212849047aca363619a16470da6c06b0d620968","afterSHA256":"1c12d6c2d1f8cd98252583075212849047aca363619a16470da6c06b0d620968","reason":"Same source body in authored semantic page"},{"oldId":"b-d0c1c5bff6881bf9f0cd5a03","targetId":"b-d0c1c5bff6881bf9f0cd5a03","page":"reading-opening","operation":"move","beforeSHA256":"2abef8836bc8264b308f99a51ee96a2155fb5eff95c9c70505a9677db7c5cec2","afterSHA256":"2abef8836bc8264b308f99a51ee96a2155fb5eff95c9c70505a9677db7c5cec2","reason":"Same source body in authored semantic page"},{"oldId":"b-16f3d921e722d7776d3e3f06","targetId":"b-16f3d921e722d7776d3e3f06","page":"reading-opening","operation":"rewrite","beforeSHA256":"1c9dca60e949866fe3d2c1e845c73d1a487da3b37a8b825f9336a144a1d3c544","afterSHA256":"749503cc42b305dbf612f9950e053a8bd470fd67b6b03c07fab171fa53f4a3cd","reason":"Same source identity with figure semantics or reading context enriched"},{"oldId":"b-001f4c08c89859fcac4d4d04","targetId":"b-001f4c08c89859fcac4d4d04","page":"reading-opening","operation":"move","beforeSHA256":"7901720f8cf1d01174de7937f30f30409cc15448d8c9d52d834f9f5d39f015da","afterSHA256":"7901720f8cf1d01174de7937f30f30409cc15448d8c9d52d834f9f5d39f015da","reason":"Same source body in authored semantic page"},{"oldId":"b-b0b7504629c23d66db24a493","targetId":"b-b0b7504629c23d66db24a493","page":"reading-opening","operation":"move","beforeSHA256":"c10e7af18ab5805c53068a7fef8219df87144cfc145a5fffab03b20d3f2c9304","afterSHA256":"c10e7af18ab5805c53068a7fef8219df87144cfc145a5fffab03b20d3f2c9304","reason":"Same source body in authored semantic page"},{"oldId":"b-2c455dd2dd67e8bd5f5b7747","targetId":"b-2c455dd2dd67e8bd5f5b7747","page":"reading-opening","operation":"move","beforeSHA256":"2b9ccef7e1220e2250cceff1abd8bdd736b42f8da5a223f26b029f44b962e807","afterSHA256":"2b9ccef7e1220e2250cceff1abd8bdd736b42f8da5a223f26b029f44b962e807","reason":"Same source body in authored semantic page"}],"blocks":[{"id":"b-84341872df1d64d15894f412","type":"h1","scope":"mask-iou-worked-example","url":"https://codingai-lec03.pages.dev/course/blocks/mask-iou-worked-example.html#b-84341872df1d64d15894f412","content":"# 0.667 是前景 IoU，0.75 才是本例两类 mIoU","sha256":"7c9539c84882d62891054298b4186066b7eed88e53712edb4ab393fee68e5242","chapter":{"id":"II","title":"区域、像素与实例"},"unit":{"id":"pixel-prediction-objective","title":"怎样直接学习一张像素类别地图？"},"page":{"id":"reading-opening","title":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","number":1},"sectionTitle":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","currentQuestion":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":null,"nextBlock":"b-d009c369797d872bc29b9268","prerequisites":[{"id":"b-d009c369797d872bc29b9268","url":"#b-d009c369797d872bc29b9268","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-d009c369797d872bc29b9268","type":"p","scope":"mask-iou-worked-example","url":"https://codingai-lec03.pages.dev/course/blocks/mask-iou-worked-example.html#b-d009c369797d872bc29b9268","content":"比较增加输出格子与增加输入证据。最近邻和双线性只规定怎样利用现有数值；离散标签与连续 logits 的插值又不同。8×8 算例清楚给出前景 IoU、背景 IoU、mIoU 和像素准确率，避免同一个小数换四个名字。","sha256":"1c12d6c2d1f8cd98252583075212849047aca363619a16470da6c06b0d620968","chapter":{"id":"II","title":"区域、像素与实例"},"unit":{"id":"pixel-prediction-objective","title":"怎样直接学习一张像素类别地图？"},"page":{"id":"reading-opening","title":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","number":1},"sectionTitle":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","currentQuestion":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-84341872df1d64d15894f412","nextBlock":"b-d0c1c5bff6881bf9f0cd5a03","prerequisites":[{"id":"b-84341872df1d64d15894f412","url":"#b-84341872df1d64d15894f412","role":"section-title"}],"relatedBlocks":[]},{"id":"b-d0c1c5bff6881bf9f0cd5a03","type":"p","scope":"mask-iou-worked-example","url":"https://codingai-lec03.pages.dev/course/blocks/mask-iou-worked-example.html#b-d0c1c5bff6881bf9f0cd5a03","content":"已知 8×8 真值 mask：前景 16、背景 48。平均降到 4×4，再最近邻还原并以 ≥.5 阈值化后，预测前景 24，其中 16 个正确、8 个误把背景当成前景；没有漏掉真前景。因此 TP=16、FP=8、FN=0、TN=40。","sha256":"2abef8836bc8264b308f99a51ee96a2155fb5eff95c9c70505a9677db7c5cec2","chapter":{"id":"II","title":"区域、像素与实例"},"unit":{"id":"pixel-prediction-objective","title":"怎样直接学习一张像素类别地图？"},"page":{"id":"reading-opening","title":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","number":1},"sectionTitle":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","currentQuestion":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-d009c369797d872bc29b9268","nextBlock":"b-16f3d921e722d7776d3e3f06","prerequisites":[{"id":"b-84341872df1d64d15894f412","url":"#b-84341872df1d64d15894f412","role":"section-title"},{"id":"b-d009c369797d872bc29b9268","url":"#b-d009c369797d872bc29b9268","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-16f3d921e722d7776d3e3f06","type":"table","scope":"mask-iou-worked-example","url":"https://codingai-lec03.pages.dev/course/blocks/mask-iou-worked-example.html#b-16f3d921e722d7776d3e3f06","content":"| 口径 | 分子与分母 | 结果 |\n| --- | --- | --- |\n| 前景 IoU | 16/(16+8+0) | 2/3 |\n| 背景 IoU | 40/(40+0+8) | 5/6 |\n| 两类 mIoU，含背景 | (2/3+5/6)/2 | .75 |\n| 像素准确率 | (16+40)/64 | .875 |","sha256":"749503cc42b305dbf612f9950e053a8bd470fd67b6b03c07fab171fa53f4a3cd","chapter":{"id":"II","title":"区域、像素与实例"},"unit":{"id":"pixel-prediction-objective","title":"怎样直接学习一张像素类别地图？"},"page":{"id":"reading-opening","title":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","number":1},"sectionTitle":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","currentQuestion":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"table":{"headers":["口径","分子与分母","结果"],"rows":[["口径","分子与分母","结果"],["前景 IoU","16/(16+8+0)","2/3"],["背景 IoU","40/(40+0+8)","5/6"],["两类 mIoU，含背景","(2/3+5/6)/2",".75"],["像素准确率","(16+40)/64",".875"]]},"previousBlock":"b-d0c1c5bff6881bf9f0cd5a03","nextBlock":"b-001f4c08c89859fcac4d4d04","prerequisites":[{"id":"b-84341872df1d64d15894f412","url":"#b-84341872df1d64d15894f412","role":"section-title"},{"id":"b-d009c369797d872bc29b9268","url":"#b-d009c369797d872bc29b9268","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-001f4c08c89859fcac4d4d04","type":"p","scope":"mask-iou-worked-example","url":"https://codingai-lec03.pages.dev/course/blocks/mask-iou-worked-example.html#b-001f4c08c89859fcac4d4d04","content":"这个过程的输入就是已知 mask，因此它隔离的是采样与阈值，不是看 RGB 图像做分割。当前 LAB11 已将名称修正，并保留最近邻与双线性两个分支。双线性分支的两类 mIoU 为 .77564，前景 IoU 仍为 2/3；平滑改变的是错误的位置和类别组成，不是自动保证每项指标都变好。","sha256":"7901720f8cf1d01174de7937f30f30409cc15448d8c9d52d834f9f5d39f015da","chapter":{"id":"II","title":"区域、像素与实例"},"unit":{"id":"pixel-prediction-objective","title":"怎样直接学习一张像素类别地图？"},"page":{"id":"reading-opening","title":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","number":1},"sectionTitle":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","currentQuestion":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-16f3d921e722d7776d3e3f06","nextBlock":"b-b0b7504629c23d66db24a493","prerequisites":[{"id":"b-84341872df1d64d15894f412","url":"#b-84341872df1d64d15894f412","role":"section-title"},{"id":"b-d009c369797d872bc29b9268","url":"#b-d009c369797d872bc29b9268","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-b0b7504629c23d66db24a493","type":"p","scope":"mask-iou-worked-example","url":"https://codingai-lec03.pages.dev/course/blocks/mask-iou-worked-example.html#b-b0b7504629c23d66db24a493","content":"读一项 mIoU 时至少应知道平均哪些类、是否含背景、空并集怎样处理，以及先累计整份混淆矩阵还是逐图计算再平均。这些口径会改变数字；具体口径比多显示小数位更重要。","sha256":"c10e7af18ab5805c53068a7fef8219df87144cfc145a5fffab03b20d3f2c9304","chapter":{"id":"II","title":"区域、像素与实例"},"unit":{"id":"pixel-prediction-objective","title":"怎样直接学习一张像素类别地图？"},"page":{"id":"reading-opening","title":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","number":1},"sectionTitle":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","currentQuestion":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-001f4c08c89859fcac4d4d04","nextBlock":"b-2c455dd2dd67e8bd5f5b7747","prerequisites":[{"id":"b-84341872df1d64d15894f412","url":"#b-84341872df1d64d15894f412","role":"section-title"},{"id":"b-d009c369797d872bc29b9268","url":"#b-d009c369797d872bc29b9268","role":"problem-context"}],"relatedBlocks":[]},{"id":"b-2c455dd2dd67e8bd5f5b7747","type":"h2","scope":"mask-iou-worked-example","url":"https://codingai-lec03.pages.dev/course/blocks/mask-iou-worked-example.html#b-2c455dd2dd67e8bd5f5b7747","content":"## 继续阅读与实验","sha256":"2b9ccef7e1220e2250cceff1abd8bdd736b42f8da5a223f26b029f44b962e807","chapter":{"id":"II","title":"区域、像素与实例"},"unit":{"id":"pixel-prediction-objective","title":"怎样直接学习一张像素类别地图？"},"page":{"id":"reading-opening","title":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","number":1},"sectionTitle":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","currentQuestion":"0.667 是前景 IoU，0.75 才是本例两类 mIoU","sourceSlides":[],"knowledgeBlock":null,"formulas":[],"previousBlock":"b-b0b7504629c23d66db24a493","nextBlock":null,"prerequisites":[{"id":"b-84341872df1d64d15894f412","url":"#b-84341872df1d64d15894f412","role":"section-title"},{"id":"b-d009c369797d872bc29b9268","url":"#b-d009c369797d872bc29b9268","role":"problem-context"}],"relatedBlocks":[]}]}