Research Papers



Research Papers

RESEARCH PAPER / Metamaterials 2021 / Aug 2021
In this work we propose a new type of metagrating solution based on a combination of two diffraction gratings embedded inside the waveguide. We demonstrate that the proposed design has high intensity across a wide angular range and can be used as single waveguide full color combiner for AR application...
RESEARCH PAPER / SPIE Optics + Photonics 2021 / Aug 2021
In this work we propose a new type of symmetrical surface relief diffractive grating for waveguide-based Augmented Reality (AR) near-eye display system with a wide field of view (FOV). We demonstrate that by using a dual mode symmetrical in-coupling system and pupil angular tiling, we can extend the overall horizontal...
RESEARCH PAPER / Cannes Next 2021 / Jul 2021
Introduction of volumetric video creation and distribution tools, introducing MPEG MIV format, presentation of a creative collaboration with external film production company.
RESEARCH PAPER / IEEE World Haptics Conference / Jul 2021
The development of tactile screens opens new perspectives for co-located image and haptic rendering, leading to the concept of “haptic images”. They emerge from the combination of image data, rendering hardware, and haptic perception. This enables one to perceive haptic feedback while manually exploring an image. This raises nevertheless two...
RESEARCH PAPER / CLIC 2021 / Jun 2021
This paper presents an end-to-end Artificial Neural Network (ANN)-based compression framework in response to the video compression task of the Challenge for Learned Image Compression (CLIC) at CVPR 2021. In this framework, the video frames are divided into Groups Of Pictures (GOPs) in which each frame can be encoded in...
RESEARCH PAPER / IEEE ICC 2021 / Jun 2021
In cellular systems, IoT devices usually encounter a compromise between device’s reachability and its battery life; reducing the communication latency entails a reduced IoT device’s battery life. To mitigate this compromise, the lifetime of a battery powered wireless device can be significantly extended by deploying ultra-low power (ULP) receivers with...
RESEARCH PAPER / IEEE ICC / Jun 2021
An energy-efficient and scalable vector backscatter modulator architecture utilizing 1-bit delta-sigma digital-to- digital converter (DDC) based antenna load modulation is proposed. In contrast with the current state of the art (SoA) multiplexer-based M-QAM backscatter modulator that requires M antenna loads, the proposed architecture can be scaled to e.g. 256-QAM using...
RESEARCH PAPER / ICASSP 2021 / Jun 2021
Neural style transfer, allowing to apply the artistic style of one image to another, has become one of the most widely showcased computer vision applications. However, in the music audio domain, a direct analogue of the style transfer task has not yet been tackled. While style transfer methods have been...
RESEARCH PAPER / Jun 2021
ÄîThe 5G Edge Intelligence for Vertical Experimentation (5G-DIVE) project aims at demonstrating the technical merits and business value proposition of 5G technologies in two vertical pilots, namely the Industry 4.0 (I4.0) and Autonomous Drones Scout (ADS) pilots. This paper presents an overview of the overall 5G-DIVE solution and reports the...
RESEARCH PAPER / IEEE ICC’21 - Workshop WS10 Intelligent Surfaces, 14-23 June 2021, Montreal / Jun 2021
Intelligent re-configurable surfaces (IRS) can be used to control the propagation of electromagnetic (EM) waves. Since the IRS is not amplifying the signal it reflects, its required size can be large. This means that the near-field effect, e.g., the curvature of the propagating EM wave must be considered. In this...
RESEARCH PAPER / ITU Webinar on QIT for Networks – Use Cases / May 2021
This is a presentation for an ITU-T webinar on quantum network use cases. 
RESEARCH PAPER / IEEE TCSVT journal / May 2021
This paper presents an overview of the new technologies for in-loop processing and filtering in the Versatile Video Coding (VVC) standard.
RESEARCH PAPER / FMX 2021 / May 2021
Volumetric video creation and distribution has been developed for many years starting from light field capture and processing initiatives.  All technologies supporting the pipeline have matured and bring to the creative community new tools to create and edit real volumetric video scenes. In addition, distribution frameworks can now rely on...
RESEARCH PAPER / The International Conference on Learning Representations (ICLR) / May 2021
We introduce Skip-Window, a method to allow recurrent neural networks (RNNs) to trade off accuracy for computational cost during the analysis of a sequence while strictly limiting the computational cost. Similarly to existing approaches, Skip-Window extends existing RNN cells by adding a mechanism to encourage the model to process less...
RESEARCH PAPER / Phys Rev E / May 2021
We theoretically investigate the optical properties of one-dimensional non-Hermitian dispersive layered system with saturable gain and loss. We solve the non-homogeneous Helmholtz equation perturbatively, and by applying the modified transfer matrix method we obtain closed-form expressions for the reflection/transmission coefficients for the TM incident waves. The nonreciprocity of the scattering...
RESEARCH PAPER / IEEE Communications Magazine / Apr 2021
The  5thgeneration  (5G)  of  mobile  communications introduces  improvements  on  many  fronts  when  compared  to its previous generations. Besides the performance enhancements and  new  advances  in  radio  technologies,  it  also  integrates  other technological  domains,  such  as  cloud-to-things  continuum  and artificial  intelligence.  In  this  paper,  the  5G-DIVE  Elastic  EdgePlatform   (DEEP)   is   proposed  ...
RESEARCH PAPER / Conference on Computer Vision and Pattern Recognition / Apr 2021
Obtaining labeled data for machine learning tasks can be prohibitively expensive. 2 Active learning mitigates this issue by exploring the unlabeled data space and 3 prioritizing the selection of data that can best improve the model performance. A 4 common approach to active learning is to pick a small sample...
RESEARCH PAPER / Data Compression Conference (DCC) 2021 / Mar 2021
This paper provides an in-depth study of a framework consisting of combining trainable image and video codecs with machine task algorithms. The field of video coding optimization for machines is gaining traction, due to the increasing share of image and video content that is dedicated to be analyzed by machines,...
RESEARCH PAPER / PCS2021 / Mar 2021
An increasing share of captured images and videos are transmitted for storage and remote analysis by computer vision algorithms, rather than to be viewed by humans. Contrary to traditional standard codecs with engineered tools, neural network based codecs can be trained end-to-end to optimally compress images with respect to a...
RESEARCH PAPER / IEEE Conference on Computer Vision and Pattern Recognition 2021 / Feb 2021
Scene flow depicts the dynamics of a 3D scene, which is critical for various applications such as autonomous driving, robot navigation and AR/VR. Conventionally, scene flow is estimated from dense (or semi-dense) and regular video frames, e.g., RGB-D video frames. With the developments of deep neural networks (DNN) and depth...