Category: Uncategorized

Model Smoothness Can Predict Intra Domain and Out of Domain Generalization

Paper: Predicting Out-of-Domain Generalization with Local Manifold Smoothness

Code: none yet

You Only Look Once (YOLO) Object Detection Models

YOLOv1

Code

Paper: JOseph Redmon, Santosh Divvala, Ross Girshick, Ali Farhadi, “You Only Look Once: Unified, Real-Time Object Detection“, 2016

Code: https://github.com/AlexeyAB/darknet

YOLOv2

Paper: “YOLO9000:Better, Faster, Stronger“

Code: https://pjreddie.com/darknet/yolov2/

YOLOv3

Paper: YOLOv3: An Incremental Improvement

Code: https://pjreddie.com/darknet/yolo/

YOLOv4

Paper:

Code:

YOLOv5

Paper: none yet

Code: https://github.com/ultralytics/yolov5

YOLOv6

Paper: none yet

Code: https://github.com/meituan/YOLOv6

YOLOv7

Official YOLOv7 (state-of-the-art real-time detector) is more accurate and faster than:

  • – YOLOv5 by 120% FPS
  • – YOLOX by 180% FPS
  • – Dual-Swin-T by 1200% FPS
  • – ConvNext by 550% FPS
  • – SWIN-L CM-RCNN by 500% FPS
  • – PPYOLOE-X by 150% FPS

Paper: Chien-Yao Wang, Alexey Bochkovskiy, Hong-Yuan Mark Liao, YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors, 2022

Code: https://github.com/WongKinYiu/yolov7/releases

References

  • A Brief History of YOLO Object Detection Models From YOLOv1 to YOLOv5

Ethereum Merge

Update: March 24,2022

Ethereum
The Merge

Eventually the current Ethereum Mainnet will "merge" with the beacon chain proof-of-stake system.
This will mark the end of proof-of-work for Ethereum, and the full transition to proof-of-stake.
This is planned to precede the roll out of shard chains.
We formerly referred to this as "the docking."
This upgrade represents the official switch to proof-of-stake consensus. This eliminates the need for energy-intensive mining, and instead secures the network using staked ether. A truly exciting step in realizing the Ethereum vision – more scalability, security, and sustainability.

Reference: https://ethereum.org/en/upgrades/merge/

Probabilistic Machine Learning Book

Probabilistic Machine Learning: An Introduction

Referensi

Incorrect Labels in Global Wheat Head Dataset 2021

There are several incorrect labels in Global Wheat Head Dataset 2021. Some images are annotated in their original size. After that, the images are resized to final size (1024×1024 pixels). As the result, the bounding boxes are a little bit off.

Here is the list of affected images

Image LabelReason
0af5c1bc753619e4f5d504e5424d056af22954f04d50cd0d4a21682cfdd9a4dc.png Image file resized after annotation
4c9c82eeefaaa8b3b7300561820274c0ff576b47ada9239862f4a295cbdb18b7.png Image file resized after annotation
6be51c1a5132034427ecabaafa679fcac7c8f95e05a595df69401766b90d7890.png Image file resized after annotation

To correct the annotation, the coordinates must be multiplied by a correction factor and shifted a little bit downward. Here is the comparison between original annotation (green) and corrected annotation (red).

0af5c1bc753619e4f5d504e5424d056af22954f04d50cd0d4a21682cfdd9a4dc

Global Wheat Head Dataset 2021 Released

I believe that the entire dataset (including bounding box for test images) is released here: https://zenodo.org/record/5092309 1

This dataset is mentioned in the new paper “Global Wheat Head Dataset 2021 1” , in the “Code & Data Section”

I did a quick look, and I found that image filenames are changed.

Top image: from AIcrowd competition file
Bottom image: from complete dataset

Error Message from AMD Ryzen 5 5600X

A Ryzen 5 5600X processor produces the following error message:

kernel:[920812.434956] [Hardware Error]: Corrected error, no action required.
Message from syslogd@b550 at Jun  9 04:51:14 ...
 kernel:[920812.441570] [Hardware Error]: CPU:1 (19:21:0) MC15_STATUS[Over|CE|-|-|-|-|-|-|-]: 0xc04100a49b07c7cb
Message from syslogd@b550 at Jun  9 04:51:14 ...
 kernel:[920812.447096] [Hardware Error]: IPID: 0x0000000000000000
Message from syslogd@b550 at Jun  9 04:51:14 ...
 kernel:[920812.451923] [Hardware Error]: Microprocessor 5 Unit Ext. Error Code: 7, Instruction Cache Bank B ECC or parity error.
Message from syslogd@b550 at Jun  9 04:51:14 ...
 kernel:[920812.456683] [Hardware Error]: cache level: L3/GEN, tx: GEN

The computer continues working normally.

Wheat Head Detection Resources

AICrowd Global Wheat Challenge 2021 Solutions

AICrowd: Global Wheat Challenge 2021 solutions:

Kaggle Global WHeat Detection 2020 Solution

Kaggle Global Wheat Detection 2020

1st place solution

https://www.kaggle.com/c/global-wheat-detection/discussion/172418

Summary

  • Custom mosaic data augmentation
  • MixUp
  • Heavy augmentation
  • Data cleaning
  • EfficientDet
  • Faster RCNN FPN
  • Ensemble multi-scale model: Weighted-Boxes-Fusion, special thank @zfturbo
  • Test time augmentation(HorizontalFlip, VerticalFlip, Rotate90)
  • Pseudo labeling

2nd Place Solution

Link: https://www.kaggle.com/c/global-wheat-detection/discussion/175961

Repository: https://github.com/liaopeiyuan/TransferDet

3rd Place Solution

Code repository https://github.com/ufownl/global-wheat-detection

  • YOLOv3 from GluonCV
  • Use Darknet53 backbone
  • Use WBF over TTA
  • Use pseudo-labeling technique

Dataset Repository

Papers on Dataset

Papers on Solutions

Papers on object detection

Papers on Out Of Domain

Code Solutions

Yolo5 was not eligible for 2020 Wheat Head Challenge, but it can be used in 2021 Wheat Head Challenge [discussion]

YoloV5 vs EfficientDet

Articles

Other Wheat Dataset

Links

Installing CUDA Driver for Nvidia Cards in Ubuntu 20.04

Install CUDA driver for Nvidia cards in Ubuntu 20.04

Download CUDA driver from here:
https://developer.nvidia.com/cuda-downloads

Choose operating system (Linux), architecture (x86_64), distribution (Ubuntu), Version (20.04), Installer type (deb (local)).

https://developer.nvidia.com/cuda-downloads?target_os=Linux&target_arch=x86_64&target_distro=Ubuntu&target_version=2004&target_type=deblocal

Execute the following instructions:

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/cuda-ubuntu2004.pinn
sudo mv cuda-ubuntu2004.pin /etc/apt/preferences.d/cuda-repository-pin-600
wget https://developer.download.nvidia.com/compute/cuda/11.2.2/local_installers/cuda-repo-ubuntu2004-11-2-local_11.2.2-460.32.03-1_amd64.deb
sudo dpkg -i cuda-repo-ubuntu2004-11-2-local_11.2.2-460.32.03-1_amd64.deb
sudo apt-key add /var/cuda-repo-ubuntu2004-11-2-local/7fa2af80.pub
sudo apt-get updatesudo apt-get -y install cuda

Embedded Digital Control with Microcontrollers

Cem Unsalan, Duygun E. Barkana, H. Deniz Gurhan Embedded Digital Control with Microcontrollers: Implementation with C and Python (Wiley – IEEE)

About the Author

Cem Ünsalan, PhD, has over 20 years of experience working on signal processing and embedded systems. He received his doctorate from Ohio State University in 2003. He has published 23 papers in scientific journals and eight international books. 

Duygun E. Barkana, PhD, has over 16 years of experience working on control and robotic systems. She received her doctorate from Vanderbilt University in 2007. She has published 22 papers in scientific journals and six international book chapters. 

H. Deniz Gürhan is pursuing a PhD at Yeditepe University, where he received his BSc degree. He has over six years of experience working with guided microprocessors and digital signal processing. 

Table of Contents:

1. Introduction
2. Hardware to be Used in the Book
3. Software to be Used in the Book
4. Fundamentals of Digital Control
5. Conversion Between Analog and Digital Forms
6. Constructing Transfer Function of a System
7. Transfer Function Based Control System Analysis
8. Transfer Function Based Controller Design
9. State Space Based Control System Analysis
10. State Space Based Controller Design
11. Adaptive Control
12. Advanced Applications