- 综合排序
- 最热优先
- 最新优先
- 来自专栏chenjx85的技术专栏
leetcode-836-Rectangle Overlap
Two rectangles overlap if the area of their intersection is positive. To be clear, two rectangles that only touch at the corner or edges do not overlap. Given two rectangles, return whether they overlap.
70440发布于 2018-05-22 - 来自专栏chenjx85的技术专栏
leetcode-836-Rectangle Overlap
Two rectangles overlap if the area of their intersection is positive. To be clear, two rectangles that only touch at the corner or edges do not overlap. Given two rectangles, return whether they overlap.
45230发布于 2019-03-14 - 来自专栏Chris生命科学小站五年归档
R高级|cowplot包拼图（3）：overlap
强烈建议你读了前两节后，再来读次教程 R高级|利用cowplot包拼接图片（1）基本操作 R高级|利用cowplot包拼接图片（2）巧用NULL调节距离、排版 这次我们来讲overlap 3、overlap（2） overlap到这里还没有结束，因为我们还没有画《R高级|利用cowplot包拼接图片（2）巧用NULL调节距离、排版》中最后的一幅图。 这幅图的overlap并没有将x轴和y轴对齐，而是将图片缩小、然后放在右上角，所以，我们并不要align_plot对齐x轴和y轴，直接使用ggdraw和draw_plot即可。 到目前为止，我们花了3节内容，来讲解cowplot包拼图的2种模式 1、plot_gird命令，图片排版，结合NULL，可以灵活调节图片之间的距离，当NULL对应的相对宽度或相对高度为负数值，可以实现图片的overlap
1.5K20编辑于 2023-02-28 - 来自专栏星河造梦坊专栏
小功能⭐️关于Unity Collider Physics.Overlap
前言 Unity Collider的检测目前接触到的可以分为两种： OnTrigger、OnCollider Physics.Overlap 下面我们重点记录下第二种方法的应用。 Physics.Overlap **功能：**以自身形状，向外，发射设定好长度大小的射线。可获取到射线检测到的物体。 注意： a、获取到的物体都包含自身。
54410编辑于 2024-08-16 - 来自专栏计算机视觉理论及其实现
ModuleNotFoundError: No module named ‘keras_retinanet.utils.compute_overlap‘
keras_retinanet放入site-packages里面，未能解决问题参考：成功解决ModuleNotFoundError: No module named 'keras_retinanet.utils.compute_overlap 原因是需要导入的compute_overlap文件格式是.pyx结尾的，pyx文件说明：pyx文件是python的c扩展文件，不能直接被python 解释器直接调用。需要进行转化.参考?
1.1K40编辑于 2022-09-02 - 来自专栏软件研发
解决ModuleNotFoundError: No module named keras_retinanet.utils.compute_overlap
解决ModuleNotFoundError: No module named 'keras_retinanet.utils.compute_overlap'在使用Python编写机器学习项目时，我们有时会遇到各种错误 对于这个具体的错误，缺少了名为keras_retinanet.utils.compute_overlap的模块。解决方法首先，我们需要确认确实缺少了这个模块。 比如，你可以检查是否导入了compute_overlap模块，并且模块路径是否正确指向了keras_retinanet.utils.compute_overlap。 pythonCopy code# 导入所需的模块import keras_retinanetfrom keras_retinanet.utils.compute_overlap import compute_overlap 模块中导入了compute_overlap函数。
1.7K70编辑于 2023-10-31 - 来自专栏yw的数据分析
如何获得FPKMRPKM计算需要的基因长度（考虑exon之间的overlap）
这里我们跟Cufflinks的原理一致，使用总的外显子长度，并且去除过多的重叠的外显子的部分。使用R语言,输入为基因的GTF文件
3.7K20发布于 2020-03-20 - 来自专栏倾向性评分
最强的倾向性评分方法—— 重叠加权（Overlap Weighting，OW）
准确的来说，应该是重叠加权（Overlap Weighting，OW）。 为什么用OW OW是一种PS方法，旨在模拟随机临床试验（RCT）的重要属性:临床相关的目标人群、协变量平衡和精确度。 Overlap Weighting: A Propensity Score Method That Mimics Attributes of a Randomized Clinical Trial.
9.4K21发布于 2021-01-14 - 来自专栏文献分享及代码学习
单细胞代码解析-妇科癌症单细胞转录组及染色质可及性分析10
$end2 - ft.overlap$start,ft.overlap$overlap)ft.overlap$overlap <- ifelse(ft.overlap$overlapFeature == ))ft.overlap$overlap <- as.numeric(ft.overlap$overlap)hist(ft.overlap$overlap)oe.overlap <- overlappingPeaks $end2 - oe.overlap$start,oe.overlap$overlap)oe.overlap$overlap <- ifelse(oe.overlap$overlapFeature == ))oe.overlap$overlap <- as.numeric(oe.overlap$overlap)hist(oe.overlap$overlap)encode.overlap <- overlappingPeaks (encode.overlap$overlap)hist(encode.overlap$overlap)head(ft.overlap[1:3,])head(oe.overlap[1:3,])head(
77030编辑于 2022-09-01 - 来自专栏文献分享及代码学习
单细胞代码解析-妇科癌症单细胞转录组及染色质可及性分析14
$end2 - ft.overlap$start,ft.overlap$overlap)ft.overlap$overlap <- ifelse(ft.overlap$overlapFeature == "overlapStart",ft.overlap$end - ft.overlap$start2,ft.overlap$overlap)levels(factor(ft.overlap$overlap ))ft.overlap$overlap <- as.numeric(ft.overlap$overlap)hist(ft.overlap$overlap)oe.overlap <- overlappingPeaks $end2 - oe.overlap$start,oe.overlap$overlap)oe.overlap$overlap <- ifelse(oe.overlap$overlapFeature == ))oe.overlap$overlap <- as.numeric(oe.overlap$overlap)hist(oe.overlap$overlap)encode.overlap <- overlappingPeaks
65720编辑于 2022-09-05 - 来自专栏SnailTyan
Leetcode 452. Minimum Number of Arrows to Burst Balloons
points) == 0: return 0 points.sort(key=lambda p: p[0]) total = 1 overlap = points[0] for point in points: if overlap[1] < point[0]: total += 1 overlap = point else: overlap[0] = point[0] overlap[1] = min(overlap[1], point[1]) return total Version 2 class Solution: def findMinArrowShots = points[0] for point in points: if overlap[1] < point[0]: total
31910发布于 2021-02-04 - 来自专栏SimpleAI
SUFE AI Lab@KDD'23：一种基于异常得分分布的通用损失函数
相较于以往工作，Overlap loss对于不同数据场景的适用性更好，而其天然有界的数学性质保证了基于Overlap loss模型训练的稳定性。 利用trapezoidal rule去近似估计积分 我们把上述公式定义写完整，Overlap loss最终的公式如下： 整体Overlap loss ✌️ 到这里，Overlap loss的基本逻辑和理论部分就讲述完了 说完了Overlap loss是怎么来的以及方法逻辑之后，我们来看看Overlap loss有哪些优点，以及它的通用性。 首先Overlap loss有很多个比较好的性质，也暗示了基于Overlap loss的异常检测算法性能上可能会更好： ✔️Overlap loss不依赖先验超参去更新异常得分； ✔️Overlap loss Overlap loss能够有效嵌入在多个网络框架中，这点在以往AD学术文章中经常被忽略 除此之外，我们也通过在25个数据集上大量的实证研究表明Overlap loss： ⭐️对于不同网络框架而言，Overlap
59030编辑于 2023-09-01 - 来自专栏JNing的专栏
论文阅读: Soft-NMS
根据新公式易知，soft-nms对于 低overlap (注意不是低score)的bbox 保留得更好 。 落实到代码中真的就是“One line of code”： ? 作者找了一些higly-overlap objects的场景图来验证Soft-NMS的优越性： ? Thinking Soft-NMS加强了对highly-overlap objects的正确区分，同时却也削弱了对light-overlap objects的区分能力； 本质上是对overlap情形的一种 overfit，所以它只能算是对trade-off的offset； 只有在highly-overlap objects的场景下才能真正发挥作用，普通场景下并没有多少highly-overlap，所以甚至可能有反效果 ； 个人觉得Soft-NMS其实可以叫做“Overlap-NMS”。
1.9K20发布于 2018-09-27 - 来自专栏生信宝典
R语言学习 - 韦恩图
For two-set venn, the format is <-n “100, 110, 50” -l “‘a’, ‘b’”> represents (length_a, length_b, a_b_overlap , b_c_overlap, a_c_overlap, a_b_c_overlap). 100, 110, 90, 50, 40, 40, 20” -l “‘a’, ‘b’, ‘c’, ‘d’”> represents (length_a, length_b, length_c, a_b_overlap , a_c_overlap, a_d_overlap, b_c_overlap, b_d_overlap, c_d_overlap, abc_overlap, abd_overlap, acd_overlap , bcd_overlap, abcd_overlap).
2.2K70发布于 2018-02-05 - 来自专栏Listenlii的生物信息笔记
mbio: 随机采样过程会高估微生物群落的beta多样性
因此物种overlap可以根据y而不是x得到，这样就不需要知道N。 N已知时，预测的两样本overlap是： N未知时，y= px = (a1 + a2)x。 随机抽m个球出来，若m=N,三个样本的overlap为100%。但是实际上overlap取决于采样量、球的丰度分布、群落的复杂性。由于球的丰度分布相同，只有随机采样过程会给结果带来影响。 要达到一定的overlap所需要的序列数 A为2样本，B为3样本。5万条序列期望达到的overlap都高于80%。 对于2个样本，要达到90%的overlap理论需要71400条序列； 对于3个样本，要达到90%的overlap理论需要63770条序列。
67731发布于 2020-05-29 突破RAG性能瓶颈：基于动态重叠与语义结构的分块优化方法论
(text, min_size=256, max_size=1024, overlap_ratio=0.3): # 使用BERT模型检测语义边界 boundaries = detect_semantic_boundaries 查找下一个语义边界 end = find_next_boundary(start, boundaries, max_size) # 动态计算重叠区域 overlap = min(int((end-start)*overlap_ratio), max_size//3) if start > 0: chunk = text [start-overlap : end]kfc.djmfzb.mobi else: chunk = text[start:end] : min_size: 256 max_size: 1024 base_overlap: 0.25 max_overlap: 0.4 structure_weights:
30910编辑于 2025-08-08- 来自专栏小鹏的专栏
语音剪切程序
Matlab程序： function aucut(t,overlap)% t 为切割秒数，overlap 为样本重叠秒数 mkdir('classical10s');%创建保存剪切后语音的文件夹 Desktop/matlab/speechRecognition/classical10s/', a(i).name);%记录其中一个.au 文件的名称 j=fix((length(y1)/fs-overlap )/(t-overlap));%切割后音乐的份数 for k=1:j%对每首音乐进行切割并命名 y2=y1(((k-1)*t*fs-(k-1)*overlap*fs + 1):(k*t*fs-(k-1)*overlap*fs+1)); filename=strcat(str1,'_'); filename=strcat(filename
1.1K50发布于 2018-01-09 - 来自专栏云深之无迹
频域分片采样，ifft重构后，子带交叠部分是否会有失真？
总结一下 要恢复的部分 理想情况 现实失真来源 解决方案 幅度响应 子带边界无重叠 过渡带重叠导致幅度双计 overlap-add 窗平滑 相位响应 各通道群延一致 相位不对齐造成振铃 group delay in {0,0.1,0.2,0.3,0.4} (τ=0) === overlap=0.0: 0.0272, 0.0207, 0.0140, 0.0071, 0.0000, 0.0073, 0.0148 , 0.0225, 0.0305 overlap=0.1: 0.0269, 0.0205, 0.0138, 0.0070, 0.0000, 0.0072, 0.0147, 0.0223, 0.0302 overlap=0.2: 0.0265, 0.0202, 0.0137, 0.0069, 0.0000, 0.0071, 0.0145, 0.0220, 0.0298 overlap=0.3: 0.0262 , 0.0199, 0.0135, 0.0068, 0.0000, 0.0070, 0.0143, 0.0217, 0.0294 overlap=0.4: 0.0258, 0.0197, 0.0133,
12810编辑于 2026-01-07 - 来自专栏linux驱动个人学习
1. Linux-3.14.12内存管理笔记【系统启动阶段的memblock算法（1）】
list (> 1 entry * implies an overlap) */ overlap_list[overlap_entries overlap_list[i] = overlap_list[overlap_entries-1]; } list (> 1 entry * implies an overlap) */ overlap_list[overlap_entries++ overlap_list[i] = overlap_list[overlap_entries-1]; } overlap_entries ; i++) if (overlap_list[i]->type > current_type) current_type = overlap_list[i]
97620发布于 2019-09-25 - 来自专栏贾志刚-OpenCV学堂
对象检测网络中的NMS算法详解
03 NMS超参数 两个重要的参数是score阈值与overlap阈值，任何低于score阈值的BB将会被拒绝，当两个BB的IOU大于给定的overlap阈值时候，两个检测框将会被聚类分割为同一个对象检测框 Overlap阈值需要平衡精度与抑制效果： 当overlap阈值越大、proposals boxes被压制的就越少，结果就是导致大量的FP(False Positives)，进一步导致检测精度下降与丢失 (原因在于对象与背景图像之间不平衡比率，导致FP增加数目远高于TP) 当overlap阈值越大、proposals boxes被压制的就越少，结果就是导致大量的FP(False Positives)， 当overlap阈值很小的时候，导致proposals boxes被压制的很厉害，导致recall大幅下降。 当overlap阈值很小的时候，导致proposals boxes被压制的很厉害，导致recall大幅下降。
1.5K30发布于 2019-04-29
腾讯云开发者
Copyright © 2013 - 2026 Tencent Cloud. All Rights Reserved. 腾讯云 版权所有
深圳市腾讯计算机系统有限公司 ICP备案/许可证号：粤B2-20090059 
粤公网安备44030502008569号
腾讯云计算（北京）有限责任公司 京ICP证150476号 | 京ICP备11018762号