Search Results for author: James Crowley

Found 7 papers, 1 papers with code

Composing Complex and Hybrid AI Solutions

no code implementations25 Feb 2022 Peter Schüller, João Paolo Costeira, James Crowley, Jasmin Grosinger, Félix Ingrand, Uwe Köckemann, Alessandro Saffiotti, Martin Welss

Progress in several areas of computer science has been enabled by comfortable and efficient means of experimentation, clear interfaces, and interchangable components, for example using OpenCV for computer vision or ROS for robotics.

Self-Supervised Transformers for Unsupervised Object Discovery using Normalized Cut

1 code implementation CVPR 2022 Yangtao Wang, Xi Shen, Shell Hu, Yuan Yuan, James Crowley, Dominique Vaufreydaz

For unsupervised saliency detection, we improve IoU for 4. 9%, 5. 2%, 12. 9% on ECSSD, DUTS, DUT-OMRON respectively compared to previous state of the art.

 Ranked #1 on Weakly-Supervised Object Localization on CUB-200-2011 (Top-1 Localization Accuracy metric)

object-detection Object Discovery +5

Deep learning investigation for chess player attention prediction using eye-tracking and game data

no code implementations17 Apr 2019 Justin Le Louedec, Thomas Guntz, James Crowley, Dominique Vaufreydaz

The visual attention model described in this article has been created to generate saliency maps that capture hierarchical and spatial features of chessboard, in order to predict the probability fixation for individual pixels Using a skip-layer architecture of an autoencoder, with a unified decoder, we are able to use multiscale features to predict saliency of part of the board at different scales, showing multiple relations between pieces.

Rendu basé image avec contraintes sur les gradients

no code implementations29 Dec 2018 Grégoire Nieto, Frédéric Devernay, James Crowley

In general, this works by first doing a coarse 3D reconstruction of the scene, and then using this reconstruction to establish correspondences between source and target views, followed by blending the warped views to get the final image.

3D Reconstruction

Histogram of Oriented Depth Gradients for Action Recognition

no code implementations29 Jan 2018 Nachwa Abou Bakr, James Crowley

In this paper, we report on experiments with the use of local measures for depth motion for visual action recognition from MPEG encoded RGBD video sequences.

Action Recognition Temporal Action Localization

Defining the Pose of any 3D Rigid Object and an Associated Distance

no code implementations14 Dec 2016 Romain Brégier, Frédéric Devernay, Laetitia Leyrit, James Crowley

The pose of a rigid object is usually regarded as a rigid transformation, described by a translation and a rotation.

Pose Estimation

Cannot find the paper you are looking for? You can Submit a new open access paper.