Aktivitäten pro Jahr
Abstract
Modern optical flow methods are often composed of a cascade of many independent steps or formulated as a black box neural network that is hard to interpret and analyze. In this work we seek for a plain, interpretable, but learnable solution. We propose a novel inpainting based algorithm that approaches the problem in three steps: feature selection and matching, selection of supporting points and energy based inpainting. To facilitate the inference we propose an optimization layer that allows to backpropagate through 10K iterations of a first-order method without any numerical or memory problems. Compared to recent state-of-the-art networks, our modular CNN is very lightweight and competitive with other, more involved, inpainting based methods.
Originalsprache | englisch |
---|---|
Publikationsstatus | Veröffentlicht - 4 Dez. 2018 |
Veranstaltung | 14th Asian Conference on Computer Vision: ACCV 2018 - Perth Western Australia, Perth, Australien Dauer: 4 Dez. 2018 → 6 Dez. 2018 http://accv2018.net |
Konferenz
Konferenz | 14th Asian Conference on Computer Vision |
---|---|
Kurztitel | ACCV 2018 |
Land/Gebiet | Australien |
Ort | Perth |
Zeitraum | 4/12/18 → 6/12/18 |
Internetadresse |
Fingerprint
Untersuchen Sie die Forschungsthemen von „Learning Energy Based Inpainting for Optical Flow“. Zusammen bilden sie einen einzigartigen Fingerprint.Aktivitäten
- 1 Posterpräsentation
-
Learning Energy Based Inpainting for Optical Flow
Patrick Knöbelreiter (Redner/in)
5 Dez. 2018Aktivität: Vortrag oder Präsentation › Posterpräsentation › Science to science