Bienqq
/
car-detection-bayes
2
1
Fork 0
Code Issues Pull Requests Releases Wiki Activity

updates

This commit is contained in:
Glenn Jocher 2018-08-26 11:33:36 +02:00
parent a27276f055
commit 2737b419ac
2 changed files with 30 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
README.md
View File

@ -19,7 +19,7 @@ Python 3.6 or later with the following `pip3 install -U -r requirements.txt` pac
# Training # Training
Run `train.py` to begin training after downloading COCO data with `data/get_coco_dataset.sh`. Each epoch trains on 120,000 images from the train and validate COCO sets, and tests on 5000 images from the COCO validate set. An Nvidia GTX 1080 Ti will process ~16 epochs per day. Loss plots for the bounding boxes, objectness and class confidence should appear similar to results shown here (coming soon) Run `train.py` to begin training after downloading COCO data with `data/get_coco_dataset.sh`. Each epoch trains on 120,000 images from the train and validate COCO sets, and tests on 5000 images from the COCO validate set. An Nvidia GTX 1080 Ti will process ~10 epochs/day with full augmentation, or ~15 epochs/day without input image augmentation. Loss plots for the bounding boxes, objectness and class confidence should appear similar to results shown here (coming soon)
![Alt](https://github.com/ultralytics/yolov3/blob/master/data/coco_training_loss.png "training loss") ![Alt](https://github.com/ultralytics/yolov3/blob/master/data/coco_training_loss.png "training loss")
# Inference # Inference

29
results.txt
View File

@ -3926,3 +3926,32 @@
0/998 8/416 65.3 67.7 367 353 250 652 1.75e+03 0 0 64 0 3 64 0.549 0/998 8/416 65.3 67.7 367 353 250 652 1.75e+03 0 0 64 0 3 64 0.549
0/998 9/416 65.5 68.6 359 357 254 663 1.77e+03 0 0 81 0 14 81 0.541 0/998 9/416 65.5 68.6 359 357 254 663 1.77e+03 0 0 81 0 14 81 0.541
0/998 10/416 64.6 66.9 354 349 250 650 1.73e+03 0 0 54 0 2 54 0.537 0/998 10/416 64.6 66.9 354 349 250 650 1.73e+03 0 0 54 0 2 54 0.537
0/998 0/416 73.8 77.3 570 512 319 820 2.37e+03 0 0 107 0 0 107 4
0/998 1/416 66.5 77.9 449 456 284 734 2.07e+03 0 0 79 0 0 79 0.653
0/998 2/416 64.8 71.3 397 388 266 685 1.87e+03 0 0 64 0 0 64 0.633
0/998 3/416 60.4 66.6 367 363 246 632 1.74e+03 0 0 53 0 0 53 0.654
0/998 4/416 57.3 63.1 334 332 236 604 1.63e+03 0 0 49 0 0 49 0.634
0/998 5/416 58.7 63.1 352 315 237 611 1.64e+03 0 0 81 0 0 81 0.637
0/998 6/416 55.6 61.1 327 293 228 589 1.55e+03 0 0 50 0 0 50 0.636
0/998 7/416 56.7 62.3 318 291 237 612 1.58e+03 0 0 91 0 0 91 0.647
0/998 8/416 57.2 61.7 328 291 233 603 1.57e+03 0 0 59 0 0 59 0.642
0/998 9/416 59.7 63.1 334 302 239 619 1.62e+03 0 0 81 0 0 81 0.625
0/998 10/416 59.1 61.9 325 296 236 609 1.59e+03 0 0 53 0 0 53 0.629
0/998 11/416 60.9 63.1 324 301 245 633 1.63e+03 0 0 92 0 0 92 0.648
0/998 12/416 58.7 61.2 321 289 236 612 1.58e+03 0 0 36 0 0 36 0.627
0/998 13/416 59.1 61.7 316 285 238 615 1.57e+03 0 0 78 0 0 78 0.637
0/998 14/416 59.4 61.7 309 280 239 619 1.57e+03 0 0 73 0 0 73 0.633
0/998 15/416 59.2 61.5 301 273 238 616 1.55e+03 0 0 61 0 0 61 0.657
0/998 16/416 59.6 61.3 296 268 238 615 1.54e+03 0 0 59 0 1 59 0.643
0/998 17/416 58 60 290 260 232 601 1.5e+03 0 0 38 0 0 38 0.633
0/998 18/416 60.7 63.1 295 279 242 629 1.57e+03 0 0 140 0 0 140 0.641
0/998 0/416 85 97.3 434 557 358 917 2.45e+03 0 0 113 0 0 113 3.84
0/998 1/416 71.9 84.7 392 465 303 789 2.11e+03 0 0 79 0 0 79 0.586
0/998 2/416 70.5 77.2 363 413 283 733 1.94e+03 0 0 65 0 0 65 0.569
0/998 3/416 62.7 68.6 341 388 260 674 1.8e+03 0 0 59 0 0 59 0.575
0/998 4/416 58.6 63.7 312 347 243 628 1.65e+03 0 0 46 0 0 46 0.568
0/998 5/416 60.1 64.9 312 334 242 630 1.64e+03 0 0 79 0 1 79 0.569
0/998 6/416 57.4 62.5 309 318 234 609 1.59e+03 0 0 50 0 0 50 0.563
0/998 7/416 59.6 65.4 311 323 241 629 1.63e+03 0 0 91 0 0 91 0.565
0/998 8/416 59.4 66 313 324 237 621 1.62e+03 0 0 61 0 1 61 0.561
0/998 9/416 61.3 66.7 319 333 242 636 1.66e+03 0 0 81 0 15 81 0.571