Поиск по сайту

Как оценить хорошо ли натренировался каскад? Означает ли это что все негативные примеры были "использованы" на предыдущих уровнях или же на данном уровне все негативные примеры классифицируются как негативные и классификатору больше нечему учиться? Как оценить требуемую память в пике? Вроде программа занимала ~ 1.4 Gb. p.s. выскакивает странное предупреждение, хотя папка пуста. (использовал 2.4.10) opencv_traincascade -data data/cascade/experiment1 -vec data/pos_80_80_all.vec -bg data/neg/neg.txt -numPos 98 -numNeg 10000 -numStages 10 -precalcValBufSize 1024 -precalcIdxBufSize 1024 -featureType LBP -w 80 -h 80 Training parameters are loaded from the parameter file in data folder! Please empty the data folder if you want to use your own set of parameters. PARAMETERS: cascadeDirName: data/cascade/experiment1 vecFileName: data/pos_80_80_all.vec bgFileName: data/neg/neg.txt numPos: 98 numNeg: 10000 numStages: 10 precalcValBufSize[Mb] : 1024 precalcIdxBufSize[Mb] : 1024 stageType: BOOST featureType: LBP sampleWidth: 80 sampleHeight: 80 boostType: GAB minHitRate: 0.995 maxFalseAlarmRate: 0.5 weightTrimRate: 0.95 maxDepth: 1 maxWeakCount: 100 ===== TRAINING 0-stage ===== <BEGIN POS count : consumed 98 : 98 NEG count : acceptanceRatio 10000 : 1 Precalculation time: 12.951 +----+---------+---------+ | N | HR | FA | +----+---------+---------+ | 1| 1| 1| +----+---------+---------+ | 2| 1| 1| +----+---------+---------+ | 3| 1| 1| +----+---------+---------+ | 4| 1| 0.258| +----+---------+---------+ END> Training until now has taken 0 days 1 hours 38 minutes 2 seconds. ===== TRAINING 1-stage ===== <BEGIN POS count : consumed 98 : 98 NEG count : acceptanceRatio 10000 : 0.257566 Precalculation time: 13.466 +----+---------+---------+ | N | HR | FA | +----+---------+---------+ | 1| 1| 1| +----+---------+---------+ | 2| 1| 1| +----+---------+---------+ | 3| 1| 1| +----+---------+---------+ | 4| 1| 0.2202| +----+---------+---------+ END> Training until now has taken 0 days 3 hours 23 minutes 30 seconds. ===== TRAINING 2-stage ===== <BEGIN POS count : consumed 98 : 98 NEG count : acceptanceRatio 10000 : 0.0570969 Precalculation time: 12.91 +----+---------+---------+ | N | HR | FA | +----+---------+---------+ | 1| 1| 1| +----+---------+---------+ | 2| 1| 1| +----+---------+---------+ | 3| 1| 1| +----+---------+---------+ | 4| 1| 0.1442| +----+---------+---------+ END> Training until now has taken 0 days 4 hours 56 minutes 0 seconds. ===== TRAINING 3-stage ===== <BEGIN POS count : consumed 98 : 98 NEG count : acceptanceRatio 10000 : 0.00820192 Precalculation time: 11.31 +----+---------+---------+ | N | HR | FA | +----+---------+---------+ | 1| 1| 1| +----+---------+---------+ | 2| 1| 1| +----+---------+---------+ | 3| 1| 0| +----+---------+---------+ END> Training until now has taken 0 days 6 hours 9 minutes 53 seconds. ===== TRAINING 4-stage ===== <BEGIN POS count : consumed 98 : 98 Train dataset for temp stage can not be filled. Branch training terminated.