mirror/tesseract
1
0
Fork 0
mirror of https://github.com/tesseract-ocr/tesseract.git synced 2025-01-20 07:34:28 +08:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity

Merge pull request #3657 from stweil/bcer

Browse Source 
Limit BCER to interval [0,1]
This commit is contained in:
Egor Pugin 2021-11-25 13:43:47 +03:00 committed by GitHub
commit 3313bb794b
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 4 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
src/training/unicharset/lstmtrainer.cpp
View File

@ -1324,7 +1324,7 @@ double LSTMTrainer::ComputeWinnerError(const NetworkIO &deltas) {
double LSTMTrainer::ComputeCharError(const std::vector<int> &truth_str,
const std::vector<int> &ocr_str) {
std::vector<int> label_counts(NumOutputs());
int truth_size = 0;
unsigned truth_size = 0;
for (auto ch : truth_str) {
if (ch != null_char_) {
++label_counts[ch];
@ -1336,11 +1336,12 @@ double LSTMTrainer::ComputeCharError(const std::vector<int> &truth_str,
--label_counts[ch];
}
}
int char_errors = 0;
unsigned char_errors = 0;
for (auto label_count : label_counts) {
char_errors += abs(label_count);
}
if (truth_size == 0) {
// Limit BCER to interval [0,1] and avoid division by zero.
if (truth_size <= char_errors) {
return (char_errors == 0) ? 0.0 : 1.0;
}
return static_cast<double>(char_errors) / truth_size;