Abstract

Motion blur is one of the most common artifacts in photographs and videos. Handshaken mobile cameras and motions of objects occurring during the exposure are the main cause of the blur. While sharp scenes can be captured from fast shutter speed, an aligned pair of blurry and sharp images are hard to be captured at the same time. To enable supervised learning for deblurring, we propose a way to synthesize dynamic motion blurred images from high speed cameras to construct large-scale dataset. We show deep neural networks trained on this data generalizes to real blurry images and videos. Finally, we present a high-quality REDS dataset for video deblurring and super-resolution. The REDS dataset is in high-quality in terms of the reference frames and the realism of quality degradation. The REDS dataset was employed in the NTIRE 2019 challenges on video deblurring and super-resolution.

About Seungjun Nah

Seungjun Nah is a Ph. D. student at Seoul National University, advised by Prof. Kyoung Mu Lee. He received his BS degree from Seoul National University in 2014. He has worked on computer vision research topics including deblurring, super-resolution, and neural network acceleration. He won the 1st place award from NTIRE 2017 super-resolution challenge and workshop. He co-organized the NTIRE 2019  and AIM 2019 workshops and challenges on video quality restoration. He has reviewed conference (ICCV 2019, CVPR 2018, SIGGRAPH Asia 2018) and journal (IJCV, TNNLS, TMM, TIP) paper submissions. He is one of the best reviewers in ICCV 2019. His research interests include visual quality enhancement, low-level computer vision, and efficient deep learning. He is currently a guest scientist at Max Planck Institute for Intelligent Systems.

References

[1] Seungjun Nah, Tae Hyun Kim, and Kyoung Mu Lee, "Deep Multi-scale Convolutional Neural Network for Dynamic Scene Deblurring," CVPR 2017

[2] Seungjun Nah, Sanghyun Son, and Kyoung Mu Lee, “Recurrent Neural Networks with Intra-Frame Iterations for Video Deblurring,” CVPR 2019

[3] Seungjun Nah, Sungyong Baik, Seokil Hong, Gyeongsik Moon, Sanghyun Son, Radu Timofte, and Kyoung Mu Lee, “NTIRE 2019 Challenge on Video Deblurring and Super-Resolution: Dataset and Study,” CVPRW 2019

Abstract

This work covers a new approach to 3D reconstruction. In traditional 3D reconstruction one uses multiple images of the same object to calculate a 3D model by taking information gained from the differences between the images, like camera position, illumination of the images, rotation of the object and so on, to compute a point cloud representing the object. The characteristic trait shared by all these approaches is that one can almost change everything about the image, but it is not possible to change the object itself, because one needs to find correspondences between the images. To be able to use different instances of the same object, we used a 3D DPM model that can find different parts of an object in an image, thereby detecting the correspondences between the different pictures, which we then can use to calculate the 3D model. To take this theory to practise, we gave a 3D DPM model, which was trained to detect cars, pictures of different car brands, where no pair of images showed the same vehicle and used the detected correspondences and the Factorization Method to compute the 3D point cloud. This technique leads to a completely new approach in 3D reconstruction, because changing the object itself was never done before.

Luca Rolshoven will present his BSc thesis on Thursday. 

Abstract

This presentation will focus on my bachelor thesis and the topic of facial expression recognition. While researchers achieve good results on images that were taken in laboratories under the same or at least similar conditions, the performance on more arbitrary images with different head poses and illumination is still quite poor. My presentation will focus on the latter setting and will talk about available datasets, challenges and the work that has been done so far. Moreover, I will present my model STN-COV, which is a slightly modified version of a popular neural network architecture. With this model, I was able to achieve comparable results to the current state of the art.

Collaborator Nicolas Deperrois from the Computational Neuroscience Group will give a short talk about his research tomorrow, Sep. 5, 10:00 at CVG headquarters.

Abstract

The First-Person (Egocentric) Vision paradigm allows to acquire images of the world from the user’s perspective. Compared to standard Third-Person Vision, FPV is advantageous for building intelligent wearable systems able to assist the user and augment their abilities. Given their intrinsic mobility and the ability to acquire user-related information, FPV systems need to deal with a continuously evolving environment. This poses many challenges such as localization, context and intent understanding which need to be addressed in order to deliver effective solutions.

In this talk, I will present recent research on First-Person Vision which has been carried out in the Image Processing Laboratory (IPLAB) at the University of Catania. I will first focus on some applications relying on localization and context understanding to infer the user’s behavior. Then, I’ll present EPIC-KITCHENS, a large-scale dataset for object-interaction understanding from First-Person Vision data and recent work on the task of egocentric action anticipation.

References:

http://antoninofurnari.it

http://iplab.dmi.unict.it/fpv