User Tools

Site Tools


projects:year6:6a.006.tut

6a.006.TUT - Learning (Scene & Object Recognition) from Few Examples

Project - Summary

Deep neural networks enable the color constancy algorithms to approximate the function for illuminant estimation. However, the image data required for learning the chromaticity of the illumination remains scarce. Autoencoders provide a promising paradigm to exploit the underlying structure of chromaticity of images by learning over large numbers of unlabeled Internet images, so that we can achieve a good illuminant estimation over new images. We introduce a novel color constancy algorithm by auto-encoding a large dataset of images and using the model to estimate the illumination. We use two approaches. In the first, we learn a common representation of images and then fine-tune the model to estimate the illumination and, in the second approach, we combine the two steps into one using a composite objective function to allow us to learn to reconstruct and, at the same time, regress to the illumination.

Project - Team

Team Member Role Email Phone Number Academic Site/IAB
Moncef Gabbouj PI moncef.gabbouj@tuni.fi +358 (400) 736613 Tampere University
Jenni Raitoharju Co-PI jenni.raitoharju@tut.fi +358 50 447 8418 Tampere University
Firas Laakom Researcher firas.laakom@tuni.fi 358 46 5219 0250 Tampere University
Guanqun Cao Researcher Not available Not available Tampere University
Jarno Nikkanen Project Mentor jarnon@xiaomi.com 358 50 483 5323 Intel Finland

Project - Deliverables

Deliverables
1 Dataset of Color Constancy and literature review.
2 Learning strategies for model training from few examples.
3 Evaluation of new models and comparison with existing ones.
4 Realization and integration to the final system.

Project - Documents

FilenameFilesizeLast modified
6a.006.tut_poster_ppt.pptx60.8 KiB2019/08/14 15:25
6a.006.tut_learning_scene_and_object_recognition_poster_2017_fall_meeting.pptx278.2 KiB2019/08/14 15:25
6a.006.tut_ppt_presentation.pptx60.7 KiB2019/08/14 15:25
6a.006.tut_final_report.docx755.7 KiB2019/08/14 15:25
6a.006.tut_executive_summary.pdf82.0 KiB2019/08/14 15:25
6a.006.tut_cvdi_mid-year_report.pdf114.2 KiB2019/08/14 15:25
6a.006.tut_confluence_project_page.pdf147.3 KiB2019/08/14 15:25
6a.006.tut_poster_pdf.pdf287.7 KiB2019/08/14 15:25
projects/year6/6a.006.tut.txt · Last modified: 2021/06/02 17:32 by sally.johnson