github
/
darknet
mirror of https://github.com/AlexeyAB/darknet.git
1
0
Fork 0
Code Issues Releases Wiki Activity
1877 Commits
1 Branch
7 Tags
380 MiB
Tag: Branch: Tree: a553899505
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from 'a553899505'
${ noResults }
Commit Graph

116 Commits (a553899505414ea6e8d8cb4ee7bde2a78f8f6faa)

Author SHA1 Message Date
AlexeyAB a553899505 Added Supervised and Unsupervised Contrastive loss for Classifier [net] contrastive=1 unsupervised=1 [contrastive] classes=1000
5 years ago
AlexeyAB 09991d0488 Added Contrastive loss
5 years ago
AlexeyAB 6c6f04a9b3 [net] mosaic_bound=1 fixed bboxes going out of bounds. Count rewritten bboxes. Changed self-adversarial-training.
5 years ago
AlexeyAB c625cc6503 Added [conv_lstm] time_normalizer=0.5
5 years ago
AlexeyAB e455cdef78 Added [conv_lstm] lstm_activation=tanh shortcut=1
5 years ago
AlexeyAB 269878a44d Added [conv_lstm] bottleneck=1 - to make the layer cheaper
5 years ago
AlexeyAB 73768ef6e3 Added [net] max_chart_loss=50 for scale y-axis of Loss-chart
5 years ago
AlexeyAB 88a39b31d3 Added param [yolo] resize=2.0 for resizing image [1/2 - 2x] without changing aspect ratio (while jitter=0.3 changes aspect ratio)
5 years ago
AlexeyAB f5105342be Added Gradient Centralization: [convolutional] grad_centr=1
5 years ago
AlexeyAB 424632f336 Fixed typo weights_normalizion -> weights_normalization
5 years ago
AlexeyAB ccafff912b Demo speed fixed
5 years ago
AlexeyAB 0fffb686fd Added param [yolo] objectness_smooth=1
5 years ago
AlexeyAB 6ebaef3095 Added GELU activation
5 years ago
AlexeyAB b573eab63f Minor fix
5 years ago
AlexeyAB 4786d557f9 Fixed memory leak for Training with OPENCV=1 - create pool of CPU-threads only once (bug in OpenCV - it allocates memory for each thread without deallocation)
5 years ago
AlexeyAB d4b2ed9d22 Added param train_only_bn=1 to train only Batch-normalization for this and previous layers
5 years ago
AlexeyAB 68062fa4fb Added param to [yolo] iou_thresh_kind=giou iou_thresh=0.213
5 years ago
AlexeyAB 4cd21785fc Minor fix
5 years ago
none 07e4a7d7b6 fix error quantize_image's result type.
5 years ago
AlexeyAB 6a77b8dc9a Show receptive field
5 years ago
AlexeyAB 8cb1f0c44f Minor fix
5 years ago
AlexeyAB 9919612006 Added attention during training: [net] adversarial_lr=0.05 attention=1
5 years ago
AlexeyAB 978c2e00ed self-adversarial training
5 years ago
AlexeyAB 5e904d5b1e adversarial attack - experimental
5 years ago
AlexeyAB a234a50223 Added efficientnet-lite3.cfg and activation=relu6
5 years ago
AlexeyAB 5c2ddd301e Added param to cfg: [net] gaussian_noise=40
5 years ago
AlexeyAB d11caf486d Fixed BiFPN and label_smothing for Detection
5 years ago
AlexeyAB 5e54523960 Added loss_scale for Mixed-Precision on Tensor-Cores
5 years ago
AlexeyAB c814d56ec1 Added [net] dynamic_minibatch=1 for increasing mini_batch_size when random=1 is used
5 years ago
AlexeyAB ef979a1fd2 For CPU and GPU structures in the darknet.h have the same size
5 years ago
AlexeyAB 6fb817f68b Accelerated DropBlock on GPU
5 years ago
AlexeyAB 64fb042c63 Averaging Mean/Variance for several mini_batches inside 1 batch. Path to CBN.
5 years ago
Ksenya Ryabinova f441b3025a Added -chart argument
5 years ago
vinjn 3000c3111f Expose crop_image and resize_min as LIB_API
5 years ago
AlexeyAB 298805cefe Added new [convolutional] activation=normalize_channels_softmax_maxval for ASFF
5 years ago
AlexeyAB 4b59770886 Added max_delta=10 or 1 to the [yolo] and [Gaussian_yolo] layers, to limit delta and reduce the probability of Nan.
5 years ago
AlexeyAB f1643cbf2e Added: weights_normalizion = relu or softmax for [shortcut] layer
5 years ago
AlexeyAB a66bc8d6ff Resizing Detector: random=1 (/1.4 - x1.4), random=3 (/3.0 - x3.0), random=0 (no resizing). Added param: resize_step=32 to the [net] section: use resize_step=128 for 7 subsampling layers.
5 years ago
AlexeyAB 9bd88d7fd7 weighted [shortcut] layer
5 years ago
AlexeyAB b267e34487 multilayer [shortcut] in progress
5 years ago
AlexeyAB 489ea7fdce Minor fix
5 years ago
AlexeyAB 2c08a20b22 Added flag -benchmark_layers for: classifier demo, detector test, detector train
5 years ago
AlexeyAB 114a7f942b Added command line param -dontuse_opencv for training Classifier. Also use GaussianBlur instead of bilateralFilter for blur=1 for training Classifier and Detector.
5 years ago
AlexeyAB af54d95184 Added stretch_sway
5 years ago
AlexeyAB 005714a10c Added [local_avgpool] layer
5 years ago
AlexeyAB e33ecb785e Added param counters_per_class=100,200,... <for each class> to [yolo] and [Gaussian_yolo] to solve class-imbalance-problem
6 years ago
AlexeyAB 1626f9d495 Some fixes
6 years ago
AlexeyAB 5e035565ff Rotate conv-kernel
6 years ago
AlexeyAB a08c872564 Sway conv-kernel
6 years ago
AlexeyAB 642c065c0e Fixed DropBlock [dropout] dropblock_size_abs=7 dropblock_size_rel=0.5
6 years ago