Computer Vision Problems in Plant Phenotyping, Venice, 2017
Venice, Italy, 28 October, 2017, in conjunction with ICCV 2017
The goal of this third workshop, following on from the successful CVPPP at ECCV 2014 and CVPPP at BMVC 2015 was to continue to showcase the challenges raised by and extend the state of the art in computer vision for plant phenotyping. Highlights were invited talks by B.S. Manjunath, Jiri Matas, and Bernardino Romera-Paredes, who presented different aspects of computer vision topics related to plant phenotyping.
Over the past years, this workshop series was able to steadily increase interest in computer vision for plant phenotyping, and again, this workshop was a very successful event. More than 50 people took part in the workshop, most of them scientists from the general ICCV attendees, new to the field of plant phenotyping.
Best posters of CVPPP 2017 were determined by participants voting during and after the poster session. The best poster prizes were awarded t:
- M. Pound et al., Deep Learning for Multi-task Plant Phenotyping (Best Poster Award, runner up)
- M. V. Giuffrida et al., ARIGAN: Synthetic Arabidopsis Plants Using Generative Adversarial Network (Best Poster Award, runner up)
- S. Aich and I. Stavness, Leaf Counting with Deep Convolutional and Deconvolutional Networks (Best Poster Award, main prize)
Best contribution to the Leaf Counting Challenge was:
- A. Dobrescu et al., Leveraging Multiple Datasets for Deep Leaf Counting
One of the highlights of the workshop was during the invited talk of Dr Romera-Paredes, where he emphasized that the IPPN-hosted CVPPP / PRL (Plant phenotyping dataset) can be considered the MNIST for multi-instance segmentation. In machine learning MNIST is considered THE benchmark dataset and having CVPPP being up in these ranks is considered a big win for influencing the world of computer vision with plant phenotyping problems.
For more information on the speakers, programme, sponsors, organization etc. please visit the workshops website at https://www.plant-phenotyping.org/CVPPP2017.