Welcome to the MLCAS 2019 - Sorghum head detection challenge. This challenge is a part of Second International Workshop on Machine Learning for Cyber-Agricultural Systems (MLCAS 2019). The challenge aims at identifying new approaches to advance the state-of-the-art in Sorghum head detection from UAV images. The detection of sorghum heads from crop row images will be useful in estimating yield of the crops. More details about the wokshop could be found at https://www.register.extension.iastate.edu/mlcas2019/about.
This challenge uses RGB image frames of Sorghum heads collected from an UAV. The dataset consists of 1300 images wth sorghum head annotations. You can find further information about the datasets for Training phase and Final test phase in the "Participate" tab under the heading "Get Data". The datasets could be downloaded from the "Participate" tab under the heading "Files". The training dataset contains both labelled images and unlabelled images. Participants are encouraged to use algorithms from active learning, semi-supervised learning or unsupervised learning for utilizing the unlabelled images during training.
The images in the training and validation sets are provided with annotations that indicate the bounding box for each object. The names of the ground truth text files match the names of their corresponding input images. Each line of the ground truth label file contains the classname, left top and bottom right coordinates of the bounding box. The first pixel at left top of the image is considered to be coordinate (0,0). The annotation file format is <class_name> <left> <top> <right> <bottom>. For example, the ground truth text file of he input image C1-R15-G13-DSC00677.jpeg is
sorghumHeadyieldTrail 2 68 17 88
sorghumHeadyieldTrail 3 34 21 64
sorghumHeadyieldTrail 3 404 28 442
sorghumHeadyieldTrail 4 214 28 238
sorghumHeadyieldTrail 4 268 21 292
<class_name> <confidence> <left> <top> <right> <bottom>.
sorghumHeadyieldTrail 0.769319 12 682 34 700
sorghumHeadyieldTrail 0.907736 13 972 39 999
sorghumHeadyieldTrail 0.449279 14 1069 40 1094
sorghumHeadyieldTrail 0.124237 16 1122 63 1140
sorghumHeadyieldTrail 0.535341 17 742 52 780
The top three winning teams will receive travel grants as listed below (in addition to waiver of registration fees). Students and early career researchers (within five years of receiving their terminal degree) are eligible for the awards. Note that the amounts listed below are for the whole team, not individual participants. Also, in order to be eligible for the award, at least one team member needs to attend the award ceremony on September 11 and present the team’s work on September 12 at MLCAS 2019.
1st prize $1500
2nd prize $1200
3rd prize $1000
Wei Guo (The University of Tokyo, Japan Plant Phenotyping Network, Japan)
Soumik Sarkar (Iowa State University, U.S.A.)
Asheesh Singh (Iowa State University, U.S.A.)
Baskar Ganapathysubramanian (Iowa State University, U.S.A.)
Arti Singh (Iowa State University, U.S.A.)
We evaluate model performance on the mean average precision (mAP) metric from PASCAL VOC using IoU threshold of 0.5.
The datasets are released for academic research only and it is free to researchers from educational or research institutions for non-commercial purposes. When downloading the dataset you agree not to reproduce, duplicate, copy, sell, trade, resell or exploit for any commercial purposes, any portion of the images and any portion of derived data.
Start: June 10, 2019, 11 p.m.
Description: In this phase, you can submit the result on a small sample test data and see your rank in leaderboard.
Start: July 7, 2019, midnight
Description: In this phase, you can submit the result on test dataset and see your rank in leaderboard.
July 21, 2019, midnight
You must be logged in to participate in competitions.Sign In