Giter VIP home page Giter VIP logo

c-mil's People

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar

c-mil's Issues

Unable to run train_cmil.lua

My environment is Ubuntu16.04, CUDA9.0, Cudnn7.0. When I run the training code, I run into the following problem: cannot create a directory, and cannot write files to disk

cmil.png

Error reading the data: Not sure why this file is not there

The command th train_cmil.lua 0 SSW gives the following output:
{
0 : "/home/ubuntu/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th"
1 : "0"
2 : "SSW"
-2 : "-e"
-5 : "/home/ubuntu/torch/install/bin/luajit"
-3 : "package.path="/home/ubuntu/.luarocks/share/lua/5.1/?.lua;/home/ubuntu/.luarocks/share/lua/5.1/?/init.lua;/home/ubuntu/torch/install/share/lua/5.1/?.lua;/home/ubuntu/torch/install/share/lua/5.1/?/init.lua;"..package.path; package.cpath="/home/ubuntu/.luarocks/lib/lua/5.1/?.so;/home/ubuntu/torch/install/lib/lua/5.1/?.so;"..package.cpath"
-4 : "-e"
-1 : "local k,l,_=pcall(require,"luarocks.loader") _=k and l.add_context("trepl","scm-1")"
}
SETTINGS:
{
NUM_EPOCHS : 20
DATASET : "VOC2007"
annealFromEpoch3 : true
LearningRate : 0.005
model_path : "model/CMIL.lua"
ContinuationFunc : "Log"
AnnealEpoch : 10
SUBSET_FOR_TESTING : "test"
ifContinuation : true
PROPOSALS : "SSW"
BASE_MODEL : "VGGF"
LearningRateAneal : 0.0005
RESULT_SAVE_FOLDER : "VOC2007/VGGF/CMIL-SSW"
lambda : 0.7
DetRate : 0.1
CO_TRAINING_PATTERN : "Anneal0.63"
CliqueNTOP : 200
SUBSET : "trainval"
test_epoch_num : 20
}
Pre-processing voc dataset and VGG models...
VOC has been processed
VGG models has been processed
Done
Loading Dataset in: /home/ubuntu/C-MIL-master/data/datasets/VOC2007_SSW.t7
/home/ubuntu/torch/install/bin/luajit: cannot open </home/ubuntu/C-MIL-master/data/datasets/VOC2007_SSW.t7> in mode r at /home/ubuntu/torch/pkg/torch/lib/TH/THDiskFile.c:673
stack traceback:
[C]: at 0x7f7dfc8e8460
[C]: in function 'DiskFile'
/home/ubuntu/torch/install/share/lua/5.1/torch/File.lua:405: in function 'load'
dataset.lua:8: in main chunk
[C]: in function 'dofile'
train_cmil.lua:5: in main chunk
[C]: in function 'dofile'
...untu/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
[C]: at 0x00405d50

Segmentation fault (core dumped)

When I try to run the code that's what I get
Segmentation fault (core dumped)
When I tried to debug the code, I found that the code stop when it excute
local output = self.model:forward(inputs) in function Optim:optimize(optimMethod, inputs, targets, criterion, scale) in fbnn_Optim.lua file , line 145

I anyone could help I would be forever grateful

关于论文中的凸性变化说明

你们论文中关于示例选择器损失函数在y=1时为什么会是非凸的,为什么在λ=0时改造后的函数又变为凸性的?这个性质是如何变化的?

attempt to call field 'ME_LocalConsistencyFast' (a nil value)

I have successfully compiled all things. Dataset and model are also processed.
During training, I got this error:

In 1 module of nn.ParallelTable:
In 2 module of nn.Sequential:
layers/ContinuationSubset.lua:42: attempt to call field 'ME_LocalConsistencyFast' (a nil value)
stack traceback:
        layers/ContinuationSubset.lua:42: in function 'compute_consistency_sorted'
        layers/ContinuationSubset.lua:141: in function 'ComputeIndsFast'
        layers/ContinuationSubset.lua:173: in function <layers/ContinuationSubset.lua:156>
        [C]: in function 'xpcall'
        ~/torch/install/share/lua/5.1/nn/Container.lua:63: in function 'rethrowErrors'
        ~/torch/install/share/lua/5.1/nn/Sequential.lua:44: in function <~/torch/install/share/lua/5.1/nn/Sequential.lua:41>
        [C]: in function 'xpcall'
        ~/torch/install/share/lua/5.1/nn/Container.lua:63: in function 'rethrowErrors'
        ~/torch/install/share/lua/5.1/nn/ParallelTable.lua:12: in function <~/torch/install/share/lua/5.1/nn/ParallelTable.lua:10>
        [C]: in function 'xpcall'
        ~/torch/install/share/lua/5.1/nn/Container.lua:63: in function 'rethrowErrors'
        ~/torch/install/share/lua/5.1/nn/Sequential.lua:44: in function 'forward'
        fbnn_Optim.lua:146: in function 'optimize'
        train_cmil.lua:94: in main chunk
        [C]: in function 'dofile'
        ...torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
        [C]: at 0x00405d50

It works if I exchange the comment of these two lines:
https://github.com/Winfrand/C-MIL/blob/410a66a70c131267093ed55762ad38ffe4740338/layers/ContinuationSubset.lua#L41-L42
So what's ME_LocalConsistencyFast about?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.