GET-Forschungsseminar Abstracts
Knowledge Driven Saliency for Visual Attention
Michael Knopf, GET Lab
Vortrag: Mi. 09. Februar, 17:30, Raum P 1.4.17
Zusammenfassung:
Saliency due to novelty is one of the high-level attributes that cause stimulus to visual attention system. On the other hand, saliency of a known object surrounded by unknown objects has to be experimented in computational modeling. This will be useful for detecting a change in a known environment without continuous monitoring. The task will be to connect the attention system with a memory structure to save and retrieve object features and spatial locations. This memory will be consulted to find knowledge-based saliency and perform IOR on already known objects as well as spatial locations.