Visual novelty detection with automatic scale selection
نویسندگان
چکیده
منابع مشابه
Visual novelty detection with automatic scale selection
This paper presents experiments with an autonomous inspection robot, whose task was to highlight novel features in its environment from camera images. The experiments used two different attention mechanisms — saliency map and multi-scale Harris detector — and two different novelty detection mechanisms — Grow-When-Required (GWR) neural network and an incremental Principal Component Analysis (PCA...
متن کاملVisual Attention and Novelty Detection: Experiments with Automatic Scale Selection∗
We present experiments with an autonomous inspection robot, whose task was to highlight novel features in its environment using camera images. Experiments were conducted with two different attention mechanisms — saliency map and multiscale Harris detector — and two different novelty detection mechanisms — the Grow-WhenRequired neural network and incremental PCA. For both mechanisms we compared ...
متن کاملAutomatic Memory Selection via Novelty Detection
A small video camera attached to the body of a person can record essentially the same view as the person’s own visual system. Users of the Microsoft SenseCam exploit this fact and already the benefit of such footage for memory therapy has been demonstrated [2]. Undoubtedly, these systems would profit from the automatic extraction of memorable images. Although work exists which pursues this dire...
متن کاملEdge Detection with Automatic Scale Selection Approach to Improve Coherent Visual Attention Model
An automatic scale selection approach is developed to improve the coherent visual attention model (Le Meur, O., Le Callet, P., Barba, D., Thoreau, D., 2006. A coherent computational approach to model bottom-up visual attention. IEEE Trans. Pattern Anal. Machine Intell. 28 (5), 802-817). The new approach uses linear summation to combine the automatic scale selection attention model with the cohe...
متن کاملHierarchical Novelty Detection for Visual Object Recognition
ion* TD+LOO object* 2 4 device* Relabel 4 mountain* LOO 3 lamp* DARTS 2 appetizer plate LOO course* TD+LOO dish* Relabel nutriment* DARTS
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Robotics and Autonomous Systems
سال: 2007
ISSN: 0921-8890
DOI: 10.1016/j.robot.2007.05.012