ΠΑΝΕΠΙΣΤΗΜΙΟ ΙΩΑΝΝΙΝΩΝ - ΤΜΗΜΑ ΜΗΧΑΝΙΚΩΝ Η/Υ & ΠΛΗΡΟΦΟΡΙΚΗΣ SITE MAP | ΠΑΝΕΠΙΣΤΗΜΙΟ ΙΩΑΝΝΙΝΩΝ - ΤΜΗΜΑ ΜΗΧΑΝΙΚΩΝ Η/Υ & ΠΛΗΡΟΦΟΡΙΚΗΣ ΕΛΛΗΝΙΚΑ | ΠΑΝΕΠΙΣΤΗΜΙΟ ΙΩΑΝΝΙΝΩΝ - ΤΜΗΜΑ ΜΗΧΑΝΙΚΩΝ Η/Υ & ΠΛΗΡΟΦΟΡΙΚΗΣ ENGLISH
Παρασκευή, 09/02/2018   -   Ομιλίες
Σεμινάριο Τμήματος με τίτλο:"Looking for people in images by classification of visual attribute clusters", Χριστόφορος Νίκου

Στο πλαίσιο της διοργάνωσης των σεμιναρίων του τμήματος θα πραγματοποιηθεί την Παρασκευή 09/02/2018 και ώρα 12:00 στην αίθουσα Σεμιναρίων του Τμήματος Μηχανικών Η/Υ και Πληροφορικής, ομιλία με τίτλο "Looking for people in images by classification of visual attribute clusters". Ομιλητής θα είναι ο κ. Χριστόφορος Νίκου, Αναπληρωτής Καθηγητής του Τμήματος Μηχανικών Η/Υ & Πληροφορικής, Πανεπιστημίου Ιωαννίνων.

ΠΕΡΙΛΗΨΗ
When we are interested in providing a description of an object or a human, we tend to use visual attributes to accomplish this task. For example, a laptop can have a wide screen, a silver color, and a brand logo, whereas a human can be tall, female, wearing a blue t-shirt and carrying a backpack. Visual attributes in computer vision are equivalent to the adjectives in our speech. We rely on visual attributes since: (i) they enhance our understanding by creating an image in our mind of what this object or human looks like; (ii) they narrow down the possible related results when we want to search for a product online or when we want to provide a suspect description; (iii) they can be composed in different ways to create descriptions; (iv) they generalize well as with some fine-tuning they can be applied to recognize objects for different tasks; and (v) they are a meaningful semantic representation of objects or humans that can be understood by both computers and humans. However, effectively predicting the corresponding visual attributes of a human given an image remains a challenging task. In real-life scenarios, images might be of low-resolution, humans might be partially occluded in cluttered scenes, or there might be significant pose variations.
In this talk, a deep learning method to solve multiple binary classification tasks will be presented in the above framework (e.g. “Provide the images showing a woman carrying a backpack and standing in front of a coffee-shop”). The method performs end-to-end learning by feeding a convolutional neural network with the image of a human. It will be demonstrated how both multi-task and curriculum learning may be exploited in that framework by transferring the knowledge about tasks that are relatively easy to learn to more challenging classification tasks. State-of-the-art results on publicly available datasets of humans standing with their full-body visible will be discussed.

©2010 DEPARTMENT OF COMPUTER SCIENCE & ENGINEERING - UNIVERSITY OF IOANNINA
created by data|SCIENCE

Valid XHTML 1.0 Strict Valid CSS!