Vous êtes ici : Kiosque » Annonce

Identification

Identifiant: 
Mot de passe : 

Mot de passe oublié ?
Détails d'identification oubliés ?

Annonce

30 juin 2022

Benchmark and taxonomy of methods for producing energy-aware images


Catégorie : Stagiaire


Duration: 6 months
Location: Rennes, France
Start time: ASAP
Contact: olivier.lemeur@interdigital.com

 

About InterDigital

InterDigital develops mobile and video technologies that are at the core of devices, networks, and services worldwide. We solve many of the industry’s most critical and complex technical challenges, inventing solutions for more efficient broadband networks, better video delivery, and richer multimedia experiences years ahead of market deployment. InterDigital has licenses and strategic relationships with many of the world’s leading technology companies. Founded in 1972, InterDigital is listed on NASDAQ and is included in the S&P MidCap 400® index.

Job Summary

Among all electronic media devices, the display panel is considered as the primary power consumer, requiring more than half the total energy consumed by a device. This internship aims to assess existing approaches for producing energy-aware images. Such images should generate less power consumption when displayed onscreen. In the meantime, the quality of experience (QoE), i.e. the overall quality perceived by users, has to be the same as the QoE of original images.

This internship is part of the project EAM (Energy-Aware Media) recently launched at Interdigital. In this project, we aim at designing new technologies to reduce the power consumption while maintaining the QoE. These technologies span from video compression, distribution, pre/post-processing and pre-analysis. This internship investigates existing energy-aware image processing approaches and aims to reproduce existing results in order to define a benchmark. This benchmark requires the definition of a friendly environment to assess existing methods, to setup a database in order to carry out fair
comparisons. Metrics and methods will be defined as well.

The intern will perform the following tasks while applying a research methodology:
1/ select, read papers and propose an overview of state-of-the-art energy-aware image processing algorithms
2/ select the most influential approaches and reproduce results
3/ develop and/or apply those algorithms on our own image dataset:
• contribute to setup the image dataset
• evaluate and report on quality with objective metrics, and if possible subjective tests in our laboratory
• evaluate and report on algorithmic complexity and possibly on energy consumption
4/ propose and develop improvements of a new energy-aware image method. The constraint on complexity and/or energy consumption will be integrated as well.
5/ if applicable, participate in the writing of a patent and of a scientific publication.
6/ as an additional point, if time allows, the intern will evaluate the existing methods on different electronic devices, such as smartphones.

The expected outcome of the internship is:
1/ A benchmark and toolbox of existing approaches
2/ A description of a new approach which might lead to a publication or patent
3/ A demo which demonstrates the approach on a platform to be defined

Skills:

The ideal candidate will have a research background in signal processing and image processing. A research background in machine learning would be appreciated. Knowledge of deep learning techniques, python and one of the popular deep learning frameworks (Pytorch, or similar) would be appreciated.

Keywords: Computer vision, energy-aware method, energy consumption

References:

[1] Yin, Jia-Li, et al. "Deep Battery Saver: End-to-End Learning for Power Constrained Contrast Enhancement." IEEE Transactions on Multimedia 23 (2020): 1049-1059.
[2] Shin, Yong Goo, et al. "Unsupervised deep power saving and contrast enhancement for oled displays." Conference proceedings: Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference. Kluwer Academic Publishers, 2019.
[3] Kang, Suk-Ju. "Image-quality-based power control technique for organic light emitting diode displays." Journal of Display Technology 11.1 (2015): 104-109.
[4] Lin, Chun-Han, Chih-Kai Kang, and Pi-Cheng Hsiu. "Catch your attention: Quality-retaining power saving on mobile OLED displays." 2014 51st ACM/EDAC/IEEE Design Automation Conference (DAC). IEEE, 2014.

 

Duration: 6 months

Location: Rennes, France

Start time: ASAP

Contact: olivier.lemeur@interdigital.com

 

Dans cette rubrique

(c) GdR 720 ISIS - CNRS - 2011-2022.