Publication

Learning Pose Invariant and Covariant Classifiers from Image Sequences

Mustafa Özuysal
2010
Theses

Résumé

Object tracking and detection over a wide range of viewpoints is a long-standing problem in Computer Vision. Despite significant advance in wide-baseline sparse interest point matching and development of robust dense feature models, it remains a largely open problem. Moreover, abundance of low cost mobile platforms and novel application areas, such as real-time Augmented Reality, constantly push the performance limits of existing methods. There is a need to modify and adapt these to meet more stringent speed and capacity requirements. In this thesis, we aim to overcome the difficulties due to the multi-view nature of the object detection task. We significantly improve upon existing statistical keypoint matching algorithms to perform fast and robust recognition of image patches independently of object pose. We demonstrate this on various 2D and 3D datasets. The statistical keypoint matching approaches require massive amounts of training data covering a wide range of viewpoints. We have developed a weakly supervised algorithm to greatly simplify their training for 3D objects. We also integrate this algorithm in a 3D tracking-by-detection system to perform real-time Augmented Reality. Finally, we extend the use of a large training set with smooth viewpoint variation to category-level object detection. We introduce a new dataset with continuous pose annotations which we use to train pose estimators for objects of a single category. By using these estimators' output to select pose specific classifiers, our framework can simultaneously localize objects in an image and recover their pose. These decoupled pose estimation and classification steps yield improved detection rates. Overall, we rely on image and video sequences to train classifiers that can either operate independently of the object pose or recover the pose parameters explicitly. We show that in both cases our approaches mitigate the effects of viewpoint changes and improve the recognition performance.

Source officielle

https://infoscience.epfl.ch/entities/publication/6b0db844-8fbd-4251-a38d-e7c99d5743f3

Mustafa Özuysal
2010
Theses

Résumé

Source officielle

https://infoscience.epfl.ch/entities/publication/6b0db844-8fbd-4251-a38d-e7c99d5743f3

Proximité ontologique

Information engineering

Intelligence artificielle: Vision par ordinateur

Concepts associés (32)

Object co-segmentation

In computer vision, object co-segmentation is a special case of , which is defined as jointly segmenting semantically similar objects in multiple images or video frames. It is often challenging to extract segmentation masks of a target/object from a noisy collection of images or video frames, which involves object discovery coupled with . A noisy collection implies that the object/target is present sporadically in a set of images or the object/target disappears intermittently throughout the video of interest.

Détection de contours

En et en vision par ordinateur, on appelle détection de contours les procédés permettant de repérer les points d'une qui correspondent à un changement brutal de l'intensité lumineuse. Ces changements de propriétés de l' indiquent en général des éléments importants de structure dans l'objet représenté. Ces éléments incluent des discontinuités dans la profondeur, dans l'orientation d'une surface, dans les propriétés d'un matériau et dans l'éclairage d'une scène.

Système de reconnaissance faciale

Un système de reconnaissance faciale est une application logicielle visant à reconnaître automatiquement une personne grâce à son visage. Il s'agit d'un sujet particulièrement étudié en vision par ordinateur, avec de très nombreuses publications et brevets, et des conférences spécialisées. La reconnaissance de visage a de nombreuses applications en vidéosurveillance, biométrie, robotique, indexation d'images et de vidéos, , etc. Ces systèmes sont généralement utilisés à des fins de sécurité pour déverrouiller ordinateur/mobile/console, mais aussi en domotique.

Afficher plus

Publications associées (28)

Fast and Future: Towards Efficient Forecasting in Video Semantic Segmentation

Evann Pierre Guy Courdier

Deep learning has revolutionized the field of computer vision, a success largely attributable to the growing size of models, datasets, and computational power. Simultaneously, a critical pain point arises as several computer vision applications are deploye ...

EPFL2024

Enabling Uncertainty Estimation in Iterative Neural Networks

Pascal Fua, Doruk Oner, Nikita Durasov, Minh Hieu Lê

Turning pass-through network architectures into iterative ones, which use their own output as input, is a well-known approach for boosting performance. In this paper, we argue that such architectures offer an additional benefit: The convergence rate of the ...

Curran Associates2024

Aggregating Spatial and Photometric Context for Photometric Stereo

David Honzátko

Photometric stereo, a computer vision technique for estimating the 3D shape of objects through images captured under varying illumination conditions, has been a topic of research for nearly four decades. In its general formulation, photometric stereo is an ...

EPFL2024

Afficher plus