CvSVM::train_auto() run out of memory (Bug #511)

Added by Fei DUAN over 14 years ago. Updated over 14 years ago.

Status:Done Start date:
Priority:High Due date:
Assignee:- % Done:

0%
Category:ml
Target version:-
Affected version: Operating System:
Difficulty: HW Platform:
Pull request:

Description

when the dimension of the training data is above several thousand, CvSVM::train_auto() will run out of memory after several minutes.
My testing platform is Windows XP / Win7, RAM 6G

reproduce-bug.cpp - Reproduction of a similar bug (1.5 kB) Stefan Walk, 2010-08-18 02:35 pm

Associated revisions

Revision 6a6ee640
Added by Vadim Pisarevsky over 14 years ago

fixed allocation bug in SVM training (ticket #511)

Revision 7b79eaf5
Added by Andrey Kamaev about 12 years ago

Merge pull request #511 from aritzlc:master

History

#1
Updated by Stefan Walk over 14 years ago

I've attached a sample program reproducing a similar issue - it seems the procedure underestimates the block size for cvMemStorage (so it does not really run out of memory, but the error message suggests it does ...)

OpenCV Error: One of arguments' values is out of range (requested size is negative or too big) in cvMemStorageAlloc, file /home/walk/downloads/opencv/modules/core/src/datastructs.cpp, line 332
terminate called after throwing an instance of 'cv::Exception'
what(): /home/walk/downloads/opencv/modules/core/src/datastructs.cpp:332: error: (-211) requested size is negative or too big in function cvMemStorageAlloc

#2
Updated by Vadim Pisarevsky over 14 years ago

thanks for the report and for the sample!
and sorry for delay.

the bug is fixed in r4119

  • Status changed from Open to Done
  • (deleted custom field) set to fixed

Also available in: Atom PDF