Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Video Reconstruction from Events
HDR, high-speed video reconstruction from events
Keywords: image reconstruction, videography
Event cameras have a number of advantages over standard frame-based cameras. Two of them are the high-dynamic range and high temporal resolution.
In previous work, we have successfully reconstructed images from a stream of events (see: https://youtu.be/eomALySSGVU). Now, we want to take a step further and improve this pipeline to improve the overall quality of the reconstruction. Applications range from computational photography/videography to calibration of event cameras.
This project requires previous experience in machine learning as well at least one course in computer vision. During the project, you will have the opportunities to design novel deep learning architectures tailored to event-based vision and image reconstruction. Contact us for more details.
Event cameras have a number of advantages over standard frame-based cameras. Two of them are the high-dynamic range and high temporal resolution. In previous work, we have successfully reconstructed images from a stream of events (see: https://youtu.be/eomALySSGVU). Now, we want to take a step further and improve this pipeline to improve the overall quality of the reconstruction. Applications range from computational photography/videography to calibration of event cameras.
This project requires previous experience in machine learning as well at least one course in computer vision. During the project, you will have the opportunities to design novel deep learning architectures tailored to event-based vision and image reconstruction. Contact us for more details.
The goal of this project is to extract high-dynamic range, high-frame rate video from a stream of events.
The goal of this project is to extract high-dynamic range, high-frame rate video from a stream of events.