There is a newer version of the record available.

Published April 24, 2019 | Version v6
Software Open

ultralytics/yolov3: Rectangular Inference, Conv2d + Batchnorm2d Layer Fusion

  • 1. @ultralytics Ultralytics LLC
  • 2. State Grid Electric Power Research Institute
  • 3. Master's 2nd
  • 4. Liverpool John Moores University
  • 5. Scalar Research
  • 6. @silverpond

Description

This release requires PyTorch >= v1.0.0 to function properly. Please install the latest version from https://github.com/pytorch/pytorch/releases

Breaking Changes

There are no breaking changes in this release.

Bug Fixes
  • NMS now screens out nan and inf values which caused it to hang during some edge cases.
Added Functionality
  • Rectangular Inference. detect.py now automatically processes images, videos and webcam feeds using rectangular inference, letterboxing to the minimum viable 32-multiple. This speeds up inference by up to 40% on HD video: https://github.com/ultralytics/yolov3/issues/232
  • Conv2d + Batchnorm2d Fusion: detect.py now automatically fuses the Conv2d and Batchnorm2d layes in the model before running inference. This speeds up inference by about 5-10%. https://github.com/ultralytics/yolov3/issues/224
  • Hyperparameters all parameterized and grouped togethor in train.py now. Genetic Hyperparameter Evolution code added to train.py.
Performance

https://cloud.google.com/deep-learning-vm/
Machine type: n1-standard-8 (8 vCPUs, 30 GB memory)
CPU platform: Intel Skylake
GPUs: K80 ($0.198/hr), P4 ($0.279/hr), T4 ($0.353/hr), P100 ($0.493/hr), V100 ($0.803/hr)
HDD: 100 GB SSD
Dataset: COCO train 2014

GPUs batch_size batch time epoch time epoch cost <i></i> (images) (s/batch) 1 K80 16 1.43s 175min $0.58 1 P4 8 0.51s 125min $0.58 1 T4 16 0.78s 94min $0.55 1 P100 16 0.39s 48min $0.39 2 P100 32 0.48s 29min $0.47 4 P100 64 0.65s 20min $0.65 1 V100 16 0.25s 31min $0.41 2 V100 32 0.29s 18min $0.48 4 V100 64 0.41s 13min $0.70 8 V100 128 0.49s 7min $0.80 TODO (help and PR's welcome!)

Files

ultralytics/yolov3-v6.zip

Files (236.7 kB)

Name Size Download all
md5:c3bc5779e0585a5ce0b1d681cfb69f91
236.7 kB Preview Download

Additional details

Related works