SueeH

关于gen_imglist_onet的疑问

3

在第11和12步产生了4个文件 pos/part/neg/landmark;而gen_imglist_onet 将4个文件合在一起 生成了一个文件train_onet_landmark；因此一共有5个文件； 但是又说生成tfrecord的时候写需要运行四次，那这四次运行输入的是item是哪4个？ gen_imglist_onet的意义是啥呢？ @AITTSMD

finetuning on deepglint, no improvement?

19

---------- Based on open model(trained on ms1m,test acc are 99.77/98.27/98.28), I finetuning the model on Deepglint, but test acc on lfw/cfp_fp/ageddb are 99.77/96.61/98.15, show above. Besides, I collect several face...

can't find the right one using falconn v 1.3.1

4

I built a hash table with about 20,000 data, feature dimension is 256, but when using **find_k_nearest_neighbors** ,it can't find the right one; and when using [get_unique_candidates] , it may...

In the project, " Compared to the original [TensorFlow model](https://github.com/aiff22/PyNET), this implementation contains three major modifications: Instance normalization is used in PyNET's level 1. Transposed convolutional layers are replaced with...

Problem to run Kintinuous

1

Hi, I have install kintinuous on ubuntu 16.04, but when I run the command ./Kintinuous -s 7 -v ../../vocab.yml.gz -l ../dataset/loop.klg -ri -fl -od; I get the results : Point...

why different opencv?

2

Thanks for great job for line and circle detection. After run test_ED, opencv(4.5.2) & EDLIB show different run time but same results(no diff between pixel), as follows: testED.getEdgeImage() (Original) :...

Question about estimateJS

1

Thanks for great implement. the function 'estimateJS' confused me, could you please give some explain？ [https://github.com/xmba15/automatic_lidar_camera_calibration/blob/master/src/ProbabilityHandler.cpp](url)

epl单机单卡和单机多卡训练step如何理解

1

单机单卡： 启动命令：TF_CONFIG='{"cluster":{"worker":["127.0.0.1:49119"]},"task":{"type":"worker","index":0}}' CUDA_VISIBLE_DEVICES=0 bash ./scripts/train_dp.sh  单机双卡： 启动命令：TF_CONFIG='{"cluster":{"worker":["127.0.0.1:49119"]},"task":{"type":"worker","index":0}}' CUDA_VISIBLE_DEVICES=0,1 bash ./scripts/train_dp.sh  代码修改了一下：去掉了last_step限制，数据集repeat=10，将txt改为py，可执行。 [resnet_dp.txt](https://github.com/alibaba/EasyParallelLibrary/files/12671704/resnet_dp.txt) 想请教下，这个如何理解呢？每个卡分别跑了10step？

I rewrite 'bokehBlur' by python, with 'https://github.com/evanw/glfx.js/blob/master/src/filters/blur/lensblur.js' as reference. To run py, just rename 'new_bokeh_copy.txt' to 'new_bokeh_copy.py' . Since I am not familiar with objective-c and js, I wonder if...