Skip to content

Releases: AlexeyAB/darknet

YOLOv4

30 Oct 02:45
Compare
Choose a tag to compare

YOLOv4 pre-release

15 May 14:17
Compare
Choose a tag to compare

YOLOv4

YOLOv4 consists of: https://arxiv.org/abs/2004.10934

  • Backbone: CSPDarknet53
  • Neck: SPP, PAN
  • Head: YOLOv3

Download:


Accuracy (AP) / Speed (FPS):

YOLOv4 (416x416 batch=1 FP16) - 32 FPS on Jetson Xavier AGX by using tkDNN+TensorRT: https://github.com/ceccocats/tkDNN

Size Darknet FPS (avg) tkDNN TensorRT FP32 FPS tkDNN TensorRT FP16 FPS tkDNN TensorRT FP16 batch=4 FPS Speedup
320 100.6 116 202 423 4.2x
416 82.5 103 162 284 3.5x
512 69.7 91 134 206 2.9x
680 53.6 62 100 150 2.8x

comparison_gpus

Yolo v3 optimal

02 Mar 15:10
Compare
Choose a tag to compare
Yolo v3 optimal Pre-release
Pre-release

Features:

  • fusion blocks: FPN, PAN, ASFF, BiFPN
  • network modules: ResNet, CPS, SPP, RFB
  • network architecture search: CSPResNext50, CSPDarknet53, SpineNet49, EfficientNetB0, MixNet-M
  • activations: SWISH, MISH
  • other features: weighted-[shortcut], Sigmoid scaling (Scale-sensitivity), Label smoothing, Optimal hyper parameters, Dynamic mini batch size for random shapes, Squeeze-and-excitation, Grouped convolution, MixConv (grouped [route]), Elastic-module
  • data augmentation: MixUp, CutMix, Mosaic
  • losses: MSE, GIoU, CIoU, DIoU
  • detection layers: [yolo] (fixed iou_thresh), [Gaussian_yolo]
  • detection on video (sequence of frames) - layers: [crnn] (convolutional-RNN), [conv_lstm] (Convolutional LSTM)

DOI

Darknet for Windows & Linux:

18 Feb 20:54
Compare
Choose a tag to compare

Darknet for Windows & Linux:

  • object detection: Yolo v3, Yolo v2
  • classification: ResNet, Darknet, DenseNet
  • sequence prediction RNN-layers: rnn, crnn, gru, lstm
  • Tensor Cores are used

Tested for Training and Prediction.

Tested Yolo v3 for training and detection on Windows and Linux

07 Jun 22:15
1752029
Compare
Choose a tag to compare

Can be used for training and detection Yolo v3 and v2 on Windows and Linux.
Added several performance improvements.
It supports mixed-precision training/detection using Tensor Cores on GPU Volta - set CUDNN_HALF=1 in the Makefile

Tested Yolo v2 for training and detection on Windows and Linux

03 Nov 22:32
a4e1dc9
Compare
Choose a tag to compare
Yolo_v2_tested

Update Readme.md