- 博客(12)
- 资源 (1)
- 收藏
- 关注
转载 pytorch-generative-model-collections
https://github.com/znxlwm/pytorch-generative-model-collections
2017-10-29 22:06:25 703
转载 ICCV 2017:训练GAN的16个技巧,2400+星(PPT)
作者:Soumith Chintala等编译:马文【新智元导读】本文来自ICCV 2017的Talk:如何训练GAN,FAIR的研究员Soumith Chintala总结了训练GAN的16个技巧,例如输入的规范化,修改损失函数,生成器用Adam优化,使用Sofy和Noisy标签,等等。这是NIPS 2016的Soumith Chintala作的邀请演讲的修改版本,而2016年的
2017-10-29 21:35:06 14654 1
转载 GENERATIVE ADVERSARIAL NETWORKS FOR IMAGE STEGANOGRAPHY
Steganography is collection of methods to hide secret information (“payload”) within non-secret information (“container”). Its counterpart, Steganalysis, is the practice of determining if a message co
2017-10-27 22:15:45 1328
转载 PyTorch参数初始化方法
torch.nn.inittorch.nn.init.calculate_gain(nonlinearity, param=None)[source]Return the recommended gain value for the given nonlinearity function. The values are as follows:nonlin
2017-10-26 20:32:53 4005
转载 cs231n - assignment1 - softmax 梯度推导
Softmax exerciseComplete and hand in this completed worksheet (including its outputs and any supporting code outside of the worksheet) with your assignment submission. For more details see the ass
2017-10-26 15:41:15 806
转载 pooling mean max 前向和反向传播
对于mean pooling,真的是好简单:假设pooling的窗大小是2x2, 在forward的时候啊,就是在前面卷积完的输出上依次不重合的取2x2的窗平均,得到一个值就是当前mean pooling之后的值。backward的时候,把一个值分成四等分放到前面2x2的格子里面就好了。如下forward: [1 3; 2 2] -> [2] backward: [2] -> [0.5
2017-10-26 15:38:30 576
转载 基于深度学习的Person Re-ID(综述)
一. 问题的提出 Person Re-ID 全称是 Person Re-Identification,又称为 行人重检测 or 行人再识别,直观上可以通过两种思路进行比对,一种是 通过 静态图像(still-image)进行特征比对,另一种是通过视频的时序特征(temporal)进行 Video Re-Id。 不管是采用 图像特征比对的方法 还是 结合时序特
2017-10-22 21:20:40 5102
转载 论文笔记-Person Re-identification Past, Present and Future
2016_Person Re-identification Past, Present and FutureLiang Zheng, Yi Yang, and Alexander G. Hauptmann这是一篇有关于Person re-ID 综述性文章。转载请附原文地址:http://blog.csdn.net/zdh2010xyz/article/details/537
2017-10-22 20:56:00 1178
转载 使用frp实现内网穿透
群晖NAS+frp发挥更大作用Posted by SunnyRx on October 21, 2016原文地址:http://www.sunnyrx.com/2016/10/21/simple-to-use-frp/该文章于2017年9月2日将frp版本从0.8.1更新到0.13.0,下文针对frp 0.13.0配置。NAS没有公网IP是一件很不方便的事情,尤其是在国内的网络环境
2017-10-19 15:54:13 4529 1
转载 Captioning Images with Diverse Objects
Captioning Images with Diverse ObjectsSubhashini Venugopalan, Lisa Anne Hendricks, Marcus Rohrbach, Raymond Mooney, Trevor Darrell, Kate Saenko(Submitted on 24 Jun 2016 (v1), last revise
2017-10-09 11:58:38 957
转载 Deep Reinforcement Learning-based Image Captioning with Embedding Reward
Deep Reinforcement Learning-based Image Captioning with Embedding RewardZhou Ren, Xiaoyu Wang, Ning Zhang, Xutao Lv, Li-Jia Li(Submitted on 12 Apr 2017)Image captioning is a challenging
2017-10-09 11:58:16 1007
转载 Semantic Compositional Networks for Visual Captioning
Semantic Compositional Networks for Visual CaptioningZhe Gan, Chuang Gan, Xiaodong He, Yunchen Pu, Kenneth Tran, Jianfeng Gao, Lawrence Carin, Li Deng(Submitted on 23 Nov 2016 (v1), last r
2017-10-09 11:55:33 1249
空空如也
TA创建的收藏夹 TA关注的收藏夹
TA关注的人