AI has revolutionized our capabilities for automatic visual perception. Among the various problems in computer vision, two prominent ones are Semantic segmentation and Object Detection. In this ICVGIP, we present two visual data challenges from two popular and important application domains. Winning teams will have the opportunity to present their solutions at ICVGIP on Dec. 20, 2020.
- Semantic Segmentation for Autonomous Driving
Computer vision based solutions are an important enabler for autonomous
driving. A crucial module for scene understanding is to obtain a fine-grained,
pixel-level labeling of the image. This challenge will utilize the first,
large-scale, public India-centric dataset, India Driving
Dataset (IDD) ,
over which solutions for semantic segmentation are invited. Participating
teams can download the training set from here to get started.
- Object Detection for Anti-Poaching Aerial Patrolling of Protected
AI has been successfully applied to various problems in the societal and
environmental domains. We consider the problem of mitigating poaching attacks
on endangered wild animals. This challenge will utilize the BIRDSAI dataset for automatic detection of animals and humans in thermal IR
aerial images collected from UAVs for augmenting anti-poaching patrols of
protected areas. Solutions for automatic object detection are invited.
Participating teams can download the training set from here.
For more details, please visit the Challenge page
Use this link
17th Nov - Registration starts, datasets available for download
10th Dec - Testing Phase
19th Dec - Results announced at ICVGIP 2020