fourye007-CSDN博客

原创 Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Loss

Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Lossinsights3DIou lossHomoscedastic noiseHeteroscedastic noise3D框的representation 重要结论...

2019-11-14 17:24:03 600

原创 M3D-RPN

M3D-RPNPost 3D–>2D Optimization下图是M3D-RPN优化旋转角的方法其优化角度的方式挺不错的,严格利用了3d框投影在图像上的8个角点会与2D框重合该优化旋转角的方式,是否适合depth估计的迭代优化,即大致流程如下:d_init = predict()d_init_1 = d_init + sigmad_init_2 = d_init - sigm...

2019-10-30 21:08:50 1422

原创 jupyte添加不同版本的python version

问题描述自己的系统里有python3.5 和python3.6, jupyter 里只有python3.5.刚好自己的一个ipynb用到了，python3.6问题解决先安装 ipykernelpython3.6 -m pip install ipykernel手动添加 python3.6找到自己kernels文件夹所在的地方cd /home/fourye/.local/share...

2019-10-27 17:24:50 305

原创 RuntimeError: Expected object of scalar type Long but got scalar type Double

问题说明最近在使用一个新的数据集的时候,只改了原先代码的数据路径,和一些简单数据处理代码,在跑的时候就报了如下的错误:Traceback (most recent call last): File "train.py", line 172, in <module> main(parser.parse_args()) File "train.py", line 75,...

2019-06-22 13:57:31 10240 4

转载亲手实践有效:停止python子线程

import threadingimport timeimport inspectimport ctypes def _async_raise(tid, exctype): """raises the exception, performs cleanup if needed""" tid = ctypes.c_long(tid) if not inspect.i...

2019-06-01 17:23:22 1285

转载 torch: pin-memory, page-lock

对CUDA架构而言，主机端的内存被分为两种，一种是可分页内存（pageable memroy）和页锁定内存（page-lock或 pinned）。可分页内存是由操作系统API malloc()在主机上分配的，页锁定内存是由CUDA函数cudaHostAlloc()在主机内存上分配的，页锁定内存的重要属性是主机的操作系统将不会对这块内存进行分页和交换操作，确保该内存始终驻留在物理内存中。GPU知道...

2019-03-21 11:50:27 722

转载 python——ubuntu下[Error 24: too many open files]的解决方法

我在使用多线程+协程爬虫抓取数据的时候，当我的协程数×线程数非常大的时候，会提示[Error 24: too many open files]和其他一系列错误。这篇博客就是专门针对这个问题的解决方案，亲测可行。为什么会报这个错误呢？明明只打开了几个文件啊？这应该是绝大多数人的第一反应。我当时遇到这个错误的时候也很懵，后来想了想，应该是我每一个协程都获取了访问这个文件的句柄，所以尽管你只把爬取的...

2019-03-08 11:27:54 767

原创 ubuntu安装gitbook常见的问题

ubuntu安装gitbook常见的问题npm和node的版本过低更新npmnpm和node的版本过低在使用此命令安装gitbook时sudo npm install -g gitbook-cli，报如下错误：/usr/local/lib/node_modules/npm/bin/npm-cli.js:84 let notifier = require('update-noti...

2019-02-16 00:24:20 553

原创 YoloV3

与之前版本的不同之处Boundingobject scores 使用logistic 回归预测bounding box中心点使用sigmoidClass Prediction使用binary cross-entropy 损失，使用sigmoid作为预测概率Proposal Overlap Problem一个scale多预测几个boxFeatur...

2018-04-16 17:18:32 813

原创分类Global-View

Fine-Grained recognitionrequire recognition of highly localized attributes of objects while being invariant to their pose and location in the imagePart-based models construct representations by loca

2017-11-20 09:58:46 481

原创 Highly Efficient Forward and Backward Propagation of Convolutional Neural Networks for Pixelwise Cla

Contributionseliminate all the redundant computation in convolution and pooling on images by introducing novel d-regularly sparse kernels.It generates exactly the same results as those by patch-by-pa

2017-11-06 16:24:24 476

原创 SQUEEZENET

Compelling AdvantagesSmaller CNNs require less communication across servers during distributed training.Smaller CNNs require less bandwidth to export a new model from the cloud to an autonomous car.

2017-11-05 10:32:17 527

原创 DenseNet--Densely Connected Convolutional Networks

Compelling advantagesalleviate the vanishing-gradient problemstrengthen feature propagationencourage feature reusesubstantially reduce the number of parameters.requires fewer parametersno need to

2017-11-03 22:16:33 398

原创 ResNet--Deep Residual Learning for Image Recognition

Key questionVanishing/exploding gradients hamper convergence from the beginning, as the network becomes more deeper.with the network depth increasing, accuracy gets saturated (which might be unsurpri

2017-11-03 20:57:57 310

原创 VGG--VERY DEEP CONVOLUTIONAL NETWORKS FOR LARGE-SCALE IMAGE RECOGNITION

Key Pointswe fix other parameters of the architecture, and steadily increase the depth of the network by adding more convolutional layers, which is feasible due to the use of very small (3 × 3) convol

2017-11-03 19:47:13 263

原创 Maxout Networks

Motivationin multiple dimensions a maxout unit can approximate arbitrary convex functionsContributionsmaxout is cross channel poolingmaxout enhances dropout’s abilities as a model averaging techniq

2017-11-03 19:35:28 408

原创 Network in Network

Key ProblemsCNN implicitly makes the assumption that the latent concepts are linearly separablethe data for the same concept often live on a nonlinear manifold, therefore the representations that cap

2017-11-03 17:17:07 253

原创 AlexNet--ImageNet Classification with Deep Convolutional Neural Networks

LRN(Local Response Normalization)applied this normalization after applying the ReLU nonlinearity in certain layersOverlapping PoolingWe generally observe during training that models with overlapping

2017-11-03 16:48:27 319

原创 Going deeper with convolutions

Key ProblemsBigger size has two disadvantages: prone to over-fitting with less labeled datacomputation consumingMethodThe fundamental way of solving both issues would be by ultimately moving from

2017-10-30 22:23:43 256

原创 DSOD: Learning Deeply Supervised Object Detectors from Scratch

Key ProblemsLimited structure design space.Learning bias As both the loss functions and the category distributions between classification and detection tasks are different, we argue that this will l

2017-10-29 22:16:32 531

原创 Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

Key Problemsthe distribution of each layer’s inputs changes during training, as the parameters of the previous layers change. This slows down the training by requiring lower learning rates and carefu

2017-10-29 09:36:08 288

原创编程之美--第二章问题解答

8. 给定一个具有n个元素的实数集、一个实数t以及一个整数k，请问你如何快速地确定该实数集是否存在一具有k个元素的子集，其中各元素的总和至多只能为t。解: 利用寻找最小k元素，找到第k小的元素，然后将其前k个数相加的和与t比较即可。寻找最小k元素代码如下：int partition(int *arr, int from, int to) { int low = from; int

2017-10-23 22:54:16 257

原创 Focal Loss for Dense Object Detection

Focal Loss for Dense Object Detection

2017-10-14 23:05:47 518

原创 Training Region-based Object Detectors with Online Hard Example Mining

OHEM: Training Region-based Object Detectors with Online Hard Example Mining

2017-10-14 21:13:26 373

转载 Transposed Convolution, Fractionally Strided Convolution or Deconvolution

点击打开链接反卷积（Deconvolution）的概念第一次出现是Zeiler在2010年发表的论文Deconvolutional networks中，但是并没有指定反卷积这个名字，反卷积这个术语正式的使用是在其之后的工作中(Adaptive deconvolutional networks for mid a

2017-10-08 15:55:44 425

原创 23. Merge k Sorted Lists

Merge k sorted linked lists and return it as one sorted list. Analyze and describe its complexity.本人的思想：就是从第一条linkedlist逐渐和下面的linkedlist合并。/** * Definition for singly-linked list. *

2017-09-10 10:54:00 211

原创 96. Unique Binary Search Trees

Given n, how many structurally unique BST's (binary search trees) that store values 1...n?For example,Given n = 3, there are a total of 5 unique BST's. 1 3 3 2 1 \

2017-08-28 09:24:49 182

原创 86. Partition List

Given a linked list and a value x, partition it such that all nodes less than x come before nodes greater than or equal to x.You should preserve the original relative order of the nodes in each of

2017-08-27 20:58:57 242

原创 389. Find the Difference

Given two strings s and t which consist of only lowercase letters.String t is generated by random shuffling string s and then add one more letter at a random position.Find the letter that

2017-08-23 23:01:52 218

原创 Add to List 167. Two Sum II - Input array is sorted

Given an array of integers that is already sorted in ascending order, find two numbers such that they add up to a specific target number.The function twoSum should return indices of the two number

2017-08-23 10:28:59 219

原创 199. Binary Tree Right Side View

Given a binary tree, imagine yourself standing on the right side of it, return the values of the nodes you can see ordered from top to bottom.For example:Given the following binary tree, 1

2017-08-23 10:26:35 165

原创 Add to List 374. Guess Number Higher or Lower

We are playing the Guess Game. The game is as follows:I pick a number from 1 to n. You have to guess which number I picked.Every time you guess wrong, I'll tell you whether the number is h

2017-08-22 22:39:31 173

原创 368. Largest Divisible Subset

Given a set of distinct positive integers, find the largest subset such that every pair (Si, Sj) of elements in this subset satisfies: Si % Sj = 0 or Sj % Si = 0.If there are multiple solution

2017-08-22 18:56:14 223

原创 367. Valid Perfect Square

Given a positive integer num, write a function which returns True if num is a perfect square else False.Note: Do not use any built-in library function such as sqrt.Example 1:Input: 16Return

2017-08-22 10:11:12 148

原创 Find All Duplicates in an Array

Given an array of integers, 1 ≤ a[i] ≤ n (n = size of array), some elements appear twice and others appear once.Find all the elements that appear twice in this array.Could you do it without ex

2017-08-22 09:50:03 180

原创 Dungeon Game

The demons had captured the princess (P) and imprisoned her in the bottom-right corner of a dungeon. The dungeon consists of M x N rooms laid out in a 2D grid. Our valiant knight (K) was initially p

2017-08-22 09:16:49 196

原创 Number of 1 Bits

Write a function that takes an unsigned integer and returns the number of ’1' bits it has (also known as the Hamming weight).For example, the 32-bit integer ’11' has binary representation 00000000

2017-08-21 09:05:52 225

原创 Intersection of Two Arrays

Given two arrays, write a function to compute their intersection.Example:Given nums1 = [1, 2, 2, 1], nums2 = [2, 2], return [2].Note:Each element in the result must be unique.The result

2017-08-20 15:39:50 223

原创 Evaluate Reverse Polish Notation

Evaluate the value of an arithmetic expression in Reverse Polish Notation.Valid operators are +, -, *, /. Each operand may be an integer or another expression.Some examples: ["2", "1",

2017-08-20 14:42:02 206

原创 Arranging Coins

You have a total of n coins that you want to form in a staircase shape, where every k-th row must have exactly k coins.Given n, find the total number of full staircase rows that can be formed.

2017-08-19 13:28:04 218

opencv缺失文件

使用Cmake生成OpenCV3.2文件时会报缺少部分文件： boostdesc_bgm_bi.i boostdesc_bgm_hd.i boostdesc_bgm.i boostdesc_binboost_256.i boostdesc_binboost_128.i boostdesc_binboost_064.i vgg_generated_120.i vgg_generated_80.i boostdesc_lbgm.i ippicv_linux_20151201.tgz vgg_generated_64.i vgg_generated_48.i protobuf-cpp-3.1.0.tar.gz

2017-07-25

空空如也

TA创建的收藏夹 TA关注的收藏夹

TA关注的人

opencv缺失文件

the kth smallest elem

斐波那契

散列表（HashMap）

斐波那契堆（Fibonacci Heap）

二项堆（Binomial Heap）

红黑树（Red Black Tree）

二叉搜索树

ViewPager欢迎界面

代码大全（第二版）Code.Complete.Second.Edition

空空如也