Challenge Breakthrough: develop an innovative algorithmic model to track objects in thermal images that will aid in decision-making. Submit your data to be scored against the challenge metrics and You could support the advancement of robotics and autonomous driving.
Challenge Overview The newest generation of autonomous vehicles are expected to successfully navigate in environments with challenging lighting conditions. This is where common, photometric visual perception-based tracking algorithms are put to their limits. We are now exploring the use of radiometric sensors, providing infrared frames (instead of photometric, RGB ones) for object tracking in low-light conditions. The challenge put forward is to develop approaches that can effectively track objects in the dark, in both structured and unstructured environments, including pedestrians, vehicles (cars, trucks, buggies, motorcycles), among others.
Challenge Teams are encouraged to leverage publicly available datasets with similar content to TII's labeled sample dataset and to utilize color-to-thermal domain adaptation techniques due to a lack of publicly available annotated thermal video sequences. Submissions will be evaluated against criteria set by TII, based on common tracking metrics in the RGB (red, green, blue) domain.
The Challenge Develop a neural network-based object classification model using the provided annotated thermal images that detects and tracks multiple instances, within the given set of object classes, at a time.
The winning Solutions will be able to:
Most Importantly (solutions should attempt to achieve these at a minimum):
- Track multiple objects in a variety of heterogeneous environments, based on thermal images,
- Robustly track multiple objects at the same time; and
- Run standalone and with few dependencies.
Link to the opportunity: https://www.herox.com/TIIInfraredTracking?utm_source=catalonia&utm_medium=website&utm_campaign=marsxr2_promo
Suport en la inscripció: valoritzacio@udg.edu