AI2FUTURE2022 / 2 days / 40+ moving & inspiring discussion and presentations
13th & 14th of October 2022, Kraš Auditorium, Zagreb, Croatia
Full Professor
Siniša Šegvić
UniZg-FER
Siniša Šegvić completed his PhD degree in 2004 at the University of Zagreb. He spent one year as a postdoc researcher at IRISA Rennes (2006) and another year as a postdoc researcher at TU Graz (2007). Currently he is a full professor at the University of Zagreb. His research expertise is in the fields of computer vision and deep learning, with special interests in natural scene understanding for autonomous vehicles and safe traffic.
He is a reviewer at top conferences of computer vision and artificial intelligence. His research group has contributed competitive submissions to dense open-set recognition benchmarks and won several challenges in semantic segmentation of road-driving scenes.
Siniša’s
Idea
Applications of generative approaches for artificial intelligence
Most machine learning applications leverage discriminative models and supervised learning. However, discriminative models are unable to generate new content or measure plausibility of the data. Additionally, they are prone to learning simple decision rules which perform well on standard benchmarks but fail in real-world scenarios. Generative approaches are a promising approach towards alleviating these problems due to ability to learn complementary information from the training dataset.
We are especially interested in a class of generative approaches that optimizes the data likelihood due to resistance to mode collapse. This presentation shall introduce prominent generative approaches and provide an overview of application fields in computer vision and natural language processing.
Siniša’s
Session
There are many sessions and workshops at the conference, please select the ones that are great for you, but please also reserve some time for networking and meeting other great people.
THINK
Deep Learning for Smart Vehicles and Safe Roads
Recent advances in deep convolutional models have caused unprecedented growth of computer vision performance. This has opened exciting applications in the fields of smart vehicles and safe roads. Pixel-level image understanding can be achieved by associating each image window with a meaningful class such as ‘road’, ‘terrain’, ‘sidewalk’ or ‘person’. The resulting semantic map reveals the kind of surface terrain in front of the vehicle, and may be used to recover the traversability map required for motion planning. Depth can be recovered by predicting a disparity field which maximizes similarity between two stereo images.
The resulting dense reconstruction can be used to detect obstacles and potholes. Classification of short video clips allows simultaneous detection of a variety of road safety attributes in video. We illustrate these opportunities by experiments on large natural images from public datasets Cityscapes and KITTI, as well as on a large video dataset collected by an industrial partner. The obtained results confirm feasibility of the presented approaches on commodity and embedded hardware.
The resulting dense reconstruction can be used to detect obstacles and potholes. Classification of short video clips allows simultaneous detection of a variety of road safety attributes in video. We illustrate these opportunities by experiments on large natural images from public datasets Cityscapes and KITTI, as well as on a large video dataset collected by an industrial partner. The obtained results confirm feasibility of the presented approaches on commodity and embedded hardware.