Project Archives | Panasonic R&D Center – Core Tech Development Hub

Company News

Panasonic ranked #1 in Mugshot category of NIST FRVT 1:1 Challenge

Dec 08, 2022 | Project

PRDCSG once again kept Panasonic’s flag flying high in the global arena.

In the latest Performance Summary of the NIST* FRVT** 1:1 Verification released on 7^th Nov 2022, we successfully clinched the No. 1 position in the Mugshot*** category.

The NIST FRVT challenge is the world’s most respected standard of facial recognition benchmarks, having attracted a total of 843 algorithms and 299 unique developers.

We have participated in this ongoing challenge by working hand in hand with Panasonic Connect since 2017. A related press release has been issued at https://news.panasonic.com/jp/press/jn221206-2 (in Japanese) and https://news.panasonic.com/global/press/en221206-3 (in English).

* NIST (National Institute of Standards and Technology) is an agency of the United States Department of Commerce.

** FRVT (Face Recognition Vendor Test) is an ongoing competition organized independently by NIST which invites vendors from all over the world to submit facial recognition algorithms.

*** Mugshot category focuses on facial images which are captured in constrained and cooperative mode, and labelled as being of type “mugshot”.

Mr. Philip Ong of Panasonic R&D Center Singapore has been awarded EnterpriseSG-SSC/SAC Commendation Award

Jul 15, 2022 | Project

Enterprise Singapore has awarded Mr. Philip Ong with EnterpriseSG-SSC/SAC Commendation Award 2022 for the contribution to the Singapore Standardization Programme in the area of Smart City.

Mr. Ong has participated in IEC SyC Smart Cities since 2017, and in Singapore co-convened the Electrical and Electricity Standards Committee TC 2 Buildings Facilities and Services WG10 City Service Continuity. Mr. Ong has provided significant contribution for the development of IEC 63152:2020 Smart Cities – City Service Continuity against disasters – The role of the electrical supply, which reached to the publication in 2020. Currently Mr. Ong is participating in the development of IEC 63152-2: Systems Reference Deliverable – City Service Continuity – Implementation Guideline and City Service cases, of which publication is forecasted in 2023.

Panasonic R&D Center Singapore continues to strive for standardization activities to develop and thrive Panasonic’s business as well as to contribute to the national and international standardization activities.

* EnterpriseSG-SSC/SAC Commendation Award is the award jointly organized by Enterprise Singapore, the Singapore Accreditation Council (SAC) and Singapore Standards Council (SSC), to show appreciation to SAC and SSC partners for the tremendous contributions to Singapore’s Standards and Accreditation programmes, in raising Singapore’s standardization and accreditation profile nationally and internationally.

Mr. Philip Ong (right) receiving the EnterpriseSG-SSC/SAC Commendation Award

Panasonic R&D Center Singapore’s Journal Paper Published in the April 2022 Issue of the IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)

May 05, 2022 | Project

Teaming up with Nanjing University of Science and Technology (NUST), Panasonic R&D Center Singapore co-authored a journal paper titled “Covariance Attention for Semantic Segmentation” which has been published in the April 2022 issue of the IEEE Transactions on Pattern Analysis and Machine Intelligence, the world’s top journal in the field of AI, machine learning, computer vision and pattern recognition. In the article, we present a low-complexity deep learning approach that exploits the covariance matrix to encode the dependencies over local and global cues of the scene. The spatial and channel covariance attention modules are designed, respectively, to boost the accuracy of semantic segmentation (assigning a label to every pixel in the image with a corresponding class). We build a new deep learning network named CANet for semantic segmentation by using the proposed covariance attention modules to achieve the state-of-the-art performance on multiple challenging datasets. This novel AI technology can be used in various practical applications such as autonomous driving, robot navigation, safety monitoring, building inspection, medical image analysis, to name a few.

Panasonic R&D Center Singapore marking the first full-year deployment of our AI-based Visitor Counting solution with the latest technology update at the Singapore Flyer’s Time Capsule

Feb 11, 2022 | Project

Panasonic R&D Center Singapore deployed our AI-based Visitor Counting solution at the iconic Singapore Flyer – Asia’s largest giant observation wheel back in January 2021, in conjunction with the then unveiling of its new attraction, Time Capsule – an immersive and multisensory attraction to rediscover the Singapore history with a time-travelling robot. As the Time Capsule celebrated its first full year of operation, we also marked the one full-year deployment of our solution by updating our AI-based technology to a newer version which offers value-added enhancement in terms of higher accuracy and more user-friendly visualization. Our Video Analytics engine has enabled automatic counting of incomings and outgoings of visitors, which can go up to thousands daily during operating days, thus saving manual labour which would otherwise be required at the three-storey site. Our enhanced solution is expected to assist Straco Leisure, which owns the Singapore Flyer, more effectively in enforcing regulatory compliance with Singapore’s COVID-19 Safe Management Measures of the day. To date, the sustained deployment of our AI solution has been a testament to customer acceptance of our digital solution making a real positive impact on their business operations.

Panasonic R&D Center Singapore ranked in 3rd place in the NIST FRVT 1:1 Verification

Jan 28, 2022 | Project

In the latest Performance Summary of the NIST* FRVT** 1:1 Verification [last updated: 24^th January 2022], Panasonic R&D Center Singapore’s December 2021 submission (“psl-009”) was ranked in 3^rd place in the Mugshot*** category, closing the gap with the 1^st– and 2^nd-ranked algorithms which were both submitted by another company, thus effectively making Panasonic R&D Center Singapore #2 company for the category on the leaderboard at this point in time. The ongoing competition is the gold standard of facial recognition benchmarks in the world, having attracted a total of 702 algorithms and 255 unique developers since year 2017. Panasonic R&D Center Singapore will strive to keep improving our rankings in future submissions for not only in this category, but also in others, to promote Panasonic’s standing on the global stage.

* NIST (National Institute of Standards and Technology) is an agency of the United States Department of Commerce.

** FRVT (Face Recognition Vendor Test) is an ongoing competition organized independently by NIST which invites vendors from all over the world to submit facial recognition algorithms.

*** Mugshot category focuses on facial images which are captured in constrained and cooperative mode, and labelled as being of type “mugshot”.

Panasonic’s Flagship LZ2000 OLED TV Continually Powered by Panasonic R&D Center Singapore’s Neural Network Architecture

Jan 21, 2022 | Project

At this year’s Consumer Electronics Show (CES) held on 4^th -7^th January, Panasonic unveiled its new flagship OLED TV for 2022, the LZ2000, which is a successor to last year’s impressive JZ2000. Like the JZ2000, the LZ2000 continues to adopt the Lightweight-and-Versatile Network (LVNet) architecture, which was co-developed by Panasonic R&D Center Singapore and the then Appliances Company of Panasonic Corporation back in FY2019, to realize its AI-based features. Same as the JZ2000, the LZ2000 is able to automatically adjust its pictures based on the type of content being displayed and the light level in the room. The LZ2000 is additionally able to factor in the ambient light temperature, thus ensuring the best picture quality possible at any given time. Its Auto AI picture mode works even more quickly than it did in last year’s model. The LZ2000 features some innovations on the audio front, too.

References:

https://na.panasonic.com/explore/press-release/panasonic-introduces-lz2000-oled-tv-2022

https://www.panasonic.com/uk/corporate/news/articles/panasonic-introduces-lz2000-its-flagship-oled-for-2022.html

Panasonic R&D Center Singapore Won a Silver Medal in a Kaggle Competition

Jan 21, 2022 | Project

Panasonic R&D Center Singapore won a silver medal in a Kaggle competition titled “Petfinder.my – Pawpularity Contest”, which started on 23^rd September 2021 and ended on 15^th January 2022, after finishing the top 1.58% out of 3,537 participating teams on the final private leaderboard. In this competition*, participants analyzed raw images and metadata to predict the popularity of shelter pet photos, by training and testing AI models on PetFinder.my’s thousands of pet profiles.

* Background of this competition [extracted from Kaggle]:

A picture is worth a thousand words. But did you know a picture can save a thousand lives? Millions of stray animals suffer on the streets or are euthanized in shelters every day around the world. You might expect pets with attractive photos to generate more interest and be adopted faster. But what makes a good picture? With the help of data science, you may be able to accurately determine a pet photo’s appeal and even suggest improvements to give these rescue animals a higher chance of loving homes.

Panasonic Auto Tracking Software Key AW-SF100/200 Powered by Panasonic R&D Center Singapore’s Face Recognition Module

Dec 28, 2021 | Project

Panasonic R&D Center Singapore’s Face Recognition module has been incorporated into the latest version of the Panasonic Auto Tracking Software Key AW-SF100/200, which was released on 24^th December 2021. In light of an increasing need for video distribution across online seminars/conferences and classes to facilitate remote working as well as temporary closure of schools, it has massively catalyzed growing demand for PTZ (pan-tilt-zoom) cameras in corporate and educational settings. Coupled with tracking technology, Panasonic R&D Center Singapore’s deep-learning-based technology not only offers an improved accuracy of facial recognition but enables Panasonic’s PTZ cameras to better capture presenters with almost no interruption, resulting in smooth tracking control of a natural-looking video that tracks the speaker. This technology development is a joint effort by Panasonic R&D Center Singapore and Panasonic CNS-MEBD* targeting the market segment of broadcast and professional AV.

* CNS-MEBD: Media Entertainment Business Division of Connected Solutions Company

Panasonic R&D Center Singapore Won the Jury Prize in the 2021 VIPriors Image Classification Challenge

Oct 15, 2021 | Project

Panasonic R&D Center Singapore was awarded the Jury Prize, while in 3^rd position, in the Image Classification track of the 2021 “Visual Inductive Priors for Data-Efficient Computer Vision” (VIPriors) challenge, held from October 11 to October 17, in conjunction with the premier International Conference on Computer Vision (ICCV) 2021, one of the world’s top-tier conferences on computer vision. The jury-based prize was awarded to the most interesting submission in each track. We joined the challenge by teaming up with Nanyang Technological University (NTU) to further sharpen our image-based object recognition competency and promote our standing globally. At that, we submitted to the VIPriors Workshop a technical paper that will be published on arXiv.

Panasonic introduces more of the 2021 OLED and LED TV line-up powered by Panasonic R&D Center Singapore’s AI engine

Apr 05, 2021 | Project

Following the debut of its flagship JZ2000 OLED TV at CES 2021, Panasonic unveiled more of its OLED and LED TV line-up for 2021.

https://www.panasonic.com/uk/corporate/news/articles/see-it-all-feel-it-all-panasonic-introduces-its-2021-tv-line-up.html

https://news.panasonic.com/jp/press/data/2021/03/jn210326-4/jn210326-4.html

https://news.panasonic.com/jp/press/data/2021/03/jn210326-5/jn210326-5.html

The new TV products adopted the AI engine named LVNet, which was co-developed by Panasonic R&D Center Singapore and the Digital Transformation Development Center of the Appliances Company of Panasonic Corporation.

In terms of OLED TVs, Panasonic has complemented the flagship JZ2000 TV with the new JZ1500, JZ1000 and JZ980 series, all of which, same with the JZ2000 TV, will feature the brand new HCX Pro AI processor to enable the new Auto AI picture mode. In essence, Auto AI helps deliver an optimum experience to viewers by accurately detecting what kind of content is being viewed – for example, cinema or sports – and then automatically adjusts the picture quality accordingly.

As for LED TVs, Panasonic has revealed the JX940 series for overseas markets, and the JX850, JX900 and JX950 series for the Japanese market, respectively. These LED TVs leverage the scene recognition algorithm to optimize the picture quality as well as the sound quality in Auto AI mode accordingly.

Panasonic’s newly launched JZ200 OLED TV powered by Panasonic R&D Center Singapore’s proprietary neural network named Lightweight-and-Versatile Network (LVNet)

Jan 29, 2021 | Project

Panasonic has announced the launch of its flagship OLED TV for 2021, that is, the JZ2000, which includes a brand-new processor called HCX Pro AI Processor. This processor enables auto-optimization of pictures and sound quality based on the type of content being viewed through the use of AI.

Such AI capability is powered by Panasonic R&D Center Singapore’s proprietary neural network named Lightweight-and-Versatile Network (LVNet), which was co-developed with the Digital Transformation Development Center (DXDC) of AP company.

This AI has been trained for an extended period of time using over a million different pieces of content to make it easier for the majority of our customers to enjoy all their favourite on-screen experiences in the best possible way, without them having to get bogged down in menus.

Panasonic R&D Center Singapore’s Paper Accepted for Publication in an Upcoming Issue of the IEEE Transactions on Intelligent Vehicles (T-IV)

Jan 08, 2021 | Project

Teaming up with Nanjing University of Science and Technology (NUST), Panasonic R&D Center Singapore co-authored a journal paper titled “Vehicle Detection and Disparity Estimation using Blended Stereo Images” which has been accepted for publication in an upcoming issue of the IEEE Transactions on Intelligent Vehicles, an international journal focusing on innovative research and application challenges in areas of intelligent vehicles, in particular, in automated vehicles in a roadway environment.

In the article, Panasonic R&D Center Singapore as the lead author (first author) and NUST as the corresponding author present a novel method for simultaneous vehicle detection and disparity estimation from the single blended multiple-view input image using the newly designed proposal-clique network. This novel method enables very competitive detection and disparity estimation accuracy to be obtained, even under circumstances where detection is performed on the blended image which is unfriendly to human vision.

Panasonic R&D Center Singapore’s Technology – LVNet for Photo Management – Incorporated into Panasonic’s New DIGA Models to Be Commercially Launched in Japan

Oct 01, 2020 | Project

New models of Panasonic’s Blu-ray recorder DIGA will be launched in Japan on 16^th October 2020. (Model Name: DMR-4T401・4CT401 / DMR-4T301・4CT301 / DMR-4T201・4CT201). These 4K DIGA models contain new AI-based features of photo classification and photo slideshow. The AI capability is powered by Panasonic R&D Center Singapore’s proprietary neural network named Lightweight-and-Versatile Network (LVNet), which was co-developed with the Digital Transformation Development Center of AP company. When photos and videos are imported into these DIGA models, the AI will automatically generate albums organized by subjects and events such as people, animals and landscapes. The photos classified there will then be summarized as a one-minute video, providing viewers with a new way to enjoy photography.

https://panasonic.jp/diga/products/4t401_4t301_4t201.html

Panasonic R&D Center Singapore’s paper accepted for Publication in an Upcoming Issue of the Transactions on Pattern Analysis and Machine Intelligence (TPAMI)

Sep 24, 2020 | Project

Teaming up with Nanjing University of Science and Technology (NUST), Panasonic R&D Center Singapore co-authored a journal paper titled “Covariance Attention for Semantic Segmentation” which has been accepted for publication in an upcoming issue of the Transactions on Pattern Analysis and Machine Intelligence, the world’s top journal in the field of computer vision and pattern recognition. In the article, we present a low complexity approach that exploits the covariance matrix to encode the dependencies over local and global cues of the scene. The spatial and channel covariance attention modules are designed, respectively, to boost the accuracy of semantic segmentation. We build a new deep learning network named CANet for scene parsing/semantic segmentation by using the proposed covariance attention modules to achieve the state-of-the-art performance on multiple challenging datasets.

Panasonic R&D Center Singapore’s paper accepted for presentation at the IEEE GLOBECOM 2020 and for publication in its proceedings

Aug 20, 2020 | Project

Panasonic R&D Center Singapore’s paper titled “Big Data Scenarios Simulator for Deep Learning Algorithm Evaluation for Autonomous Vehicles” has been accepted for presentation at the IEEE GLOBECOM 2020, December 7 – 11, in Taipei, Taiwan and for publication in its proceedings. The paper was co-authored with the University of Glasgow. In this paper, we propose a computer graphic simulator with a customized fisheye lens with the distortion factor. The system is able to generate a big training and testing dataset that mimics the real environment, and also able to evaluate Deep Learning models in accordance with the Euro NCAP 2020 standard. As an embodiment of the Digital Twin concept, the usage of such virtual-world technology is able to significantly reduce costs and speed up the process of developing autonomous vehicles in the real world.

Panasonic launched a “Panasonic Deep Tech Innovation Challenge” inviting local start-ups to leverage Panasonic’s technology & patents for business creation

Apr 20, 2020 | Event

Panasonic R&D Center Singapore has launched a Panasonic Deep Tech Innovation Challenge organized by Panasonic, ICMG and ACE. This Challenge is aimed at collaborating with local start-ups and/or providing Panasonic’s technology for start-ups to create their business ideas by leveraging Panasonic’s technology and patents and further enable such ideas to materialize. Through this Challenge, we aspire to work closely with start-ups to resolve local deep issues in the South-East Asian region.

This Challenge is part of the activity of the IP Strategy Office as well as the IP Group at Panasonic R&D Center Singapore in a bid to maximize a collaborative synergy by utilizing both IP and Technology.

Submission deadline for participation in this Challenge: 5 Jun 2020 @12PM (SGT)

Find out more and be part of the Challenge now – https://www.panasonic-dtic.com

ACE: Singapore government’s startup support organization

ICMG: Corporate Innovation Accelerator to support open innovation activities of large companies

Panasonic wins the top spot in a Future Convenience Store Challenge at IREX2019, an International Robot Exhibition 2019

Apr 01, 2020 | Project

Panasonic R&D Center Singapore (PRDCSG) joined the Nara Institute of Science and Technology, Ritsumeikan University, Panasonic’s Robotics Promotion Office, LS, CNS-IC, IS, and AP to form a NAIST-RITS-Panasonic team and compete in the Future Convenience Store Challenge at IREX2019, one of the world’s largest robot exhibitions.
https://biz.nikkan.co.jp/eve/irex/
The Future Convenience Store Challenge at IREX2019 is a stepping stone to the World Robot Summit 2020 slated for October 2020. Panasonic R&D Center Singapore is responsible for developing camera calibration technology required for the robot to localize its own position and also to precisely locate target objects for its pick-and-place tasks. Having a tough contest against the other 10 strong teams during a 3-day intensive competition, the NAIST-RITS-Panasonic team successfully landed the top spot for stocking and disposing tasks, while obtaining the 2nd position in a ranking for both customer interaction and restroom cleaning tasks.

Paper accepted for publication in the special issue of the IEEE Transactions on Circuits and Systems for Video Technology

Dec 05, 2019 | Project

Teaming up with several world-renowned video experts, Panasonic R&D Center Singapore co-authored a paper titled “General Video Coding Technology in Responses to the Joint Call for Proposals on Video Compression with Capability beyond HEVC”, which has been accepted for publication in the special issue of the IEEE Transactions on Circuits and Systems for Video Technology. The paper describes the advanced technologies used in the upcoming video coding standard, Versatile Video Coding (VVC), in which Panasonic R&D Center Singapore’s technologies that include a triangular partitioning tool are highlighted as one of the novel technologies.

Paper accepted for publication in the IEEE Transactions on Systems, Man, and Cybernetics: Systems

Dec 05, 2019 | Project

Collaborating with Nanjing University of Science and Technology (NUST), Panasonic R&D Center Singapore co-authored a paper titled “Modular Lightweight Network for Road Object Detection using a Feature Fusion Approach”, which has been accepted for publication in the IEEE Transactions on Systems, Man, and Cybernetics: Systems. The paper presents a modular lightweight Deep Learning model for detection of road objects such as cars, pedestrians and cyclists. In a situation where objects are far away from the camera and small in size, we utilize a fast and efficient network architecture, referred to as modular feature fusion detector (MFFD), to give evidence of the improvement of ThinNet.

Panasonic R&D Center Singapore (PRDCSG) participated in the ITS World Congress 2019

Dec 05, 2019 | Event

Joining Panasonic’s booth at the ITS World Congress 2019 held at Suntec City of Singapore from October 21st to October 25th, Panasonic R&D Center Singapore promoted our presence with 3 oral presentations followed by 3 posters.

 Visual Mapping and Localization for Autonomous Vehicles
 Real-Time Perception Using Lightweight AI
 Computer Graphic Simulator for Automotive AI Algorithm Development

Panasonic R&D Center Singapore Achieves No. 1 Accuracy of Face Recognition on the IJB-C Dataset

Jun 30, 2019 | Project

Panasonic R&D Center Singapore achieved the Best Accuracy of Face Recognition on the IJB-C dataset under three different protocols, namely 1:1 mixed verification, 1:N mixed identification and 1:1 covariate verification.

On top of that, a paper titled “Deep Face Recognition Model Compression via Knowledge Transfer and Distillation” was published on 3rd Jun 2019 on arXiv, which can be accessed by the following URL: https://arxiv.org/abs/1906.00619.

In the paper, we present a novel model compression approach based on a student-teacher paradigm for face-recognition applications. The proposed approach is simple and hardware-friendly, thus enabling the deployment of high-performing face-recognition deep models into resource-constraint devices.

Panasonic R&D Center Singapore Takes Second Place at the Second Edition of the AI Driving Olympics

May 15, 2019 | Project

Panasonic R&D Center Singapore finished second among the top 5 teams that made the cut from among the 50 teams in a leading international competition – that is, the second edition of the AI Driving Olympics, where Panasonic R&D Center Singapore is last year’s champion. (Reference: https://www.duckietown.org/archives/37690)

Termed “MYF” (username), Panasonic R&D Center Singapore emerged as one of the top 2 finalists and then moved forward to the more complicated challenges of Lane Following with Vehicles and Lane Following with Vehicles and Intersections. The other finalist (termed “JBRRussia”) was eventually crowned the winner of this edition.

The competition was held in conjunction with the premier International Conference on Robotics and Automation (ICRA) 2019.

Panasonic R&D Center Singapore Won the AI Driving Olympics (AI-DO) at NeurIPS (formerly known as NIPS) 2018

Dec 08, 2018 | Project

Panasonic R&D Center Singapore (PRDCSG) teamed up with the National University of Singapore (NUS) to partake in the inaugural AI Driving Olympics (AI-DO) competition, which took place at NeurIPS (formerly known as NIPS), a top-tier international conference in machine learning, in Montréal, Canada, on 8th December 2018. The Panasonic-NUS team cruised past all the competitors to secure the top spot.

Besting the other participants in the Leaderboards of Lane Following (LF) challenge and Lane Following with Vehicles (LFV) challenge utilizing the Duckietown platform, and subsequently during the live event at NeurIPS 2018, the Panasonic-NUS team was eventually crowned the Winner of the competition.

Reference: https://www.duckietown.org/archives/32095

Panasonic Tops the IROS 2018 Mobile Manipulation Hackathon Competition 2018 as First-Place Winner

Oct 18, 2018 | Event

Panasonic R&D Center Singapore teamed up with the Nanyang Technological University to take part in the IROS Mobile Manipulation Hackathon Competition held in Madrid on from October 1-5, 2018. Besting the other participants, Panasonic R&D Center Singapore and gotsecured the first placetop spot. IROS 2018 is one of the 3 largest international conferences on intelligent robots and systems.

More than 4,000 participants from 60 countries were drawn to partake in this conference where 1,200 papers were presented, beating out all the previous records in international robotics conferences in terms of the numbers of participants and papers presented.

Panasonic R&D Center Singapore Received the ISCIT 2018 Best Paper Award

Sep 26, 2018 | Event

Teaming up with King Mongkut’s University of Technology Thonburi (KMUTT) to co-author a paper titled “Image Denoising with Deep Convolutional and Multi-directional LSTM Networks under Poisson Noise Environments”, Panasonic R&D Center Singapore (PRDCSG) received the Best Paper Award at ISCIT 2018 held in Bangkok on 26-29 September 2018.

The award-winning paper proposed a novel method that is applicable to next-generation digital cameras for capturing high-quality images in all scenes.

Panasonic Tops the 2018 Nvidia AI City Challenge as First-Place Winner for Anomaly Detection

Jun 18, 2018 | Event

Panasonic R&D Center Singapore reprises the lead in Face Recognition Technology by notching up a stunning win for Track 2 (Anomaly Detection) in the 2018 Nvidia AI City Challenge .

According to the CVPR workshop paper ,

“The 2018 NVIDIA AI City Challenge (AIC18) evaluated the application of state-of-the-art computer vision and deep learning technologies to real-world traffic analysis problems, providing insights into understanding the opportunities and gaps that need to be overcome for the pervasive use of AI in traffic and public safety solutions.”

Reference links
http://cvpr2018.thecvf.com/program/workshops
http://openaccess.thecvf.com/content_cvpr_2018_workshops/papers/w3/Naphade_The_2018_NVIDIA_CVPR_2018_paper.pdf

Panasonic R&D Center Singapore participated in TechInnovation 2017

Sep 19, 2017 | Event

Panasonic R&D Center Singapore had a presence at TechInnovation 2017, a technology-industry brokerage event organized by IPI Singapore at Sands Expo & Convention Centre on 19-20 September 2017, to exhibit 12 posters featuring technologies such as artificial intelligence, autonomous driving, data analytics, wireless communications, etc. In addition, Panasonic R&D Center Singapore made a crowd-pitching presentation on “Face Recognition and Person Re-Identification”.

Joint collaborations with I2R: achieved the No. 2 position at MEC 2017 plus a paper accepted at ICMI 2017

Sep 14, 2017 | Project

In collaboration with the Institute for Infocomm Research (I2R), Panasonic R&D Center Singapore achieved the No. 2 position in the audio-visual emotion recognition sub-challenge of the Multimodal Emotion Recognition Challenge (MEC) 2017. The challenge (*1) is aimed at the comparison of multimedia processing and machine learning methods for automatic audio and visual emotion analysis.

Apart from this achievement, collaborating with I2R, Panasonic R&D Center Singapore also co-authored a paper titled “Audio-Visual Emotion Recognition using Deep Transfer Learning and Multiple Temporal Models”.

The paper, having been accepted at ACM International Conference on Multimodal Interaction (ICMI) 2017 (*2), presents the techniques used in our contribution to Emotion Recognition in the Wild (EmotiW) 2017 under the audio-video-based sub-challenge, the purpose of which is to classify the six basic emotions (angry, sad, happy, surprise, fear and disgust) and neutral.

<Supplemental information 1>

* Jointly organized by Institute of Automation, Chinese Academy of Science (CASIA), Technical Committee of Artificial Psychology and Artificial Emotion of Chinese Association for Artificial Intelligence (CAAI), Technical Committee of Human-Computer Interaction of China Computer Federation (CCF) and the Association for the Advancement of Affective Computing (AAAC)

<Supplemental information 2>

* ICMI 2017 to be held in Glasgow, Scotland, on 13-17th November 2017 is a premier international forum for multidisciplinary research on multimodal human-human and human-computer interaction, interfaces, and system development.

Paper accepted at NIPS2017 and another paper at CoR2017

Sep 06, 2017 | Project

Panasonic R&D Center Singapore, collaborating with Assistant Professor Feng Jiashi of the National University of Singapore (NUS), co-authored a paper1 titled “Dual-Agent GANs for Photorealistic and Identity Preserving Profile Face Synthesis” that has been accepted at the Conference and Workshop on Neural Information Processing Systems (NIPS) 2017, a machine learning and computational neuroscience conference. This year, NIPS, a top-rated annual conference with a 22% acceptance rate in most cases, will be held in Long Beach, USA, from December 4th through December 9th, 2017.

Moreover, Panasonic R&D Center Singapore, in collaboration with Professor David Hsu of NUS, co-authored a paper2 titled “Intention-Net: Integrated Planning and Deep Learning for Autonomous Navigation” that has been accepted at the Conference on Robot Learning (CoRL) 2017, a new annual international conference focusing on the intersection of robotics and machine learning. This year’s CoRL will be held in Mountain View, California, USA, from November 13th through November 15th, 2017. Despite being its first-ever event, CoRL has invited a lot of interest from reputable sponsors (with Panasonic being one of the Gold sponsors), famous speakers (such as Yann LeCun, a computer scientist well known for Deep Learning) as well as keen attendees in the fields of robotics, autonomous perception and control.

The publication of the two papers upholds Panasonic R&D Center Singapore’s mission to promote Panasonic’s technologies at top conferences of respective fields.

<Supplemental information>

paper¹: proposes a Dual-Agent Generative Adversarial Network (DA-GAN) model which can improve the realism of a face simulator’s output using unlabelled real faces, while preserving the identity information during realism refinement. The dual agents are specifically designed for distinguishing real vs. fake and identities simultaneously.

paper²: proposes a single unified framework that connects model-based planning and deep learning for autonomous navigation, via an interface named intention to encode where robots would go in the near future.

Acceptance of two papers by ICCV’17 Workshop on MS-Celeb-1M: Recognizing One Million Celebrities in the Real World

Aug 17, 2017 | Project

Panasonic R&D Center Singapore, collaborating with the National University of Singapore (NUS), co-authored two papers that have been accepted by ICCV (International Conference on Computer Vision*) 2017 Workshop on MS-Celeb-1M: Recognizing One Million Celebrities in the Real World, which will be held in Venice, Italy, on October 28th, 2017.

The two papers are titled: “High Performance Large Scale Face Recognition with Multi-Cognition Softmax and Feature Retrieval” and “Know You at One Glance: A Compact Vector Representation for Low-Shot Learning”, which, respectively, correspond to Challenge 1 and Challenge 2 of the MS-Celeb-1M, in which we achieved the No.1 position in both competitions.

* A premier international computer vision event

Panasonic tops MS-Celeb-1M Challenges

Jul 19, 2017 | Project

Panasonic R&D Center Singapore, in collaboration with the National University of Singapore (NUS), has achieved the No. 1 position in MS-Celeb-1M Challenge 2017 , which is organized by Microsoft, for both Challenge 1 and Challenge 2. The results are in alignment with our mission to promote Panasonic’s outstanding Face Recognition technology worldwide.

In Challenge 1, the task is to recognize one million celebrities from their face images and identify them by linking to unique entity keys in a knowledge base. This task introduces very large-scale face recognition with disambiguation.

In Challenge 2, the task is to solve problem of low-shot face recognition where the number of images available for training is limited.

(Reference links)

The general website:

http://www.msceleb.org/

The leaderboard for challenge 1:

http://www.msceleb.org/leaderboard/iccvworkshop-c1

The leaderboard for challenge 2:

http://www.msceleb.org/leaderboard/c2

Panasonic R&D Center Singapore Achieving the World’s Top Best Accuracy for Face Recognition

Apr 28, 2017 | Project

In collaboration with the National University of Singapore (NUS) and the Innovation Center of Connected Solutions of Panasonic Corporation, Panasonic R&D Center Singapore has recently developed face recognition technology that has achieved the world’s Top Best Accuracy for Face Recognition by benchmarking the dataset covering all the conditions taken in the surveillance market, the dataset of which is provided by the National Institute of Standards and Technology (NIST).

Official NIST reports: NUS-Panasonic top leaderboards of IJB-A Face Recognition challenges

On 28^th April 2017, the National Institute of Standards and Technology (NIST) of the United States Department of Commerce has officially updated results of the IJB-A Face Recognition challenges in its reports. The information including the reports can be found at:

https://www.nist.gov/programs-projects/face-challenges.

The reports show that “NUS−Panasonic−032917” tops the leaderboards of both Face Verification (1:1) and Face Identification (1:N), leading the next best result by significant margins.

The “NUS−Panasonic−032917” results were the fruits of Panasonic R&D Center Singapore’s partnership with the National University of Singapore (NUS) and Panasonic CNS-IC.

Panasonic R&D Center Singapore will continue enhancing our core competency in this area, aiming to not only maintain our lead in IJB-A Face Recognition challenge but also participate in other NIST challenges.

The ITS Summit Singapore 2017 @Marina Bay Sands

Apr 12, 2017 | Event

Panasonic R&D Center Singapore partook in a 2-day Intelligent Transportation Society (ITS) Summit Singapore 2017, which attracted more than 300 delegates, industry partners and speakers from the Asia Pacific countries. At this summit, we showcased our initiatives in autonomous automotive technologies and solutions as well as our efforts to accelerate Singapore’s Autonomous Vehicle landscape.

Not only that, our booth was visited by government and industry leaders, CTOs, and experts from the Land Transport Authority (LTA), National Research Foundation, Smart Nation Programme Office and big players like ST Kinetics, Intel, HPE, NCS, QI, etc. They were impressed with Panasonic R&D Center Singapore’s computer vision and Deep Learning capabilities, in addition to the applicability of our technologies.

Moreover, Panasonic R&D Center Singapore’s rare appearance as an exhibitor surprised the visitors pleasantly, thus sparking their interest and curiosity to have follow-up and in-depth discussions with us. In the past recent years, Panasonic R&D Center Singapore has been creating awareness of our R&D strengths in Singapore’s growing digital landscape and will further bolster our presence in the ecosystem of Autonomous Vehicles.

1-day Design Thinking Workshop @Singapore Polytechnic

Dec 21, 2016 | Event

We believe creativity is an essential skill for our members trying to make a difference in their work at Panasonic R&D Center Singapore. At this, in collaboration with LWC-S (Live Well Collaborative Singapore) which is based at Singapore Polytechnic, Panasonic R&D Center Singapore organized a one-day design thinking workshop for Panasonic R&D Center Singapore’s members on December 21, 2016. Run by LWC-S to offer Panasonic R&D Center Singapore’s members the chance to be exposed to Design Thinking, the workshop was a fun-filled one-day crash course in design thinking through which Panasonic R&D Center Singapore’s members learned how to unlock their creative potential and enhance their capacity for driving innovation within Panasonic R&D Center Singapore.

IPBC Asia 2016 @Shanghai, China

Dec 04, 2016 | Event

Panasonic R&D Center Singapore-IPG attended the IPBC Asia (held in Shanghai on December 4-6, 2016), which is one of the leading events in the IP industry. IPBC Asia attracted top IP executives of global companies to gather and take center stage to discuss issues about the global IP development with a focus on how IP owners can strategically utilize their IP assets, especially in Asia. Same as last year, this year too, Panasonic was honoured as one of the “Asia IP Elites” for successful drive in creating IP value through many IP monetization and licensing deals concluded in the past year. It is an award that heralds Panasonic’s success in leveraging the patents as a strategic business asset.

Deep Learning Summit Asia @SG 2016

Oct 20, 2016 | Event

Members from Panasonic R&D Center Singapore and various Panasonic divisions and centres in Japan attended the Deep Learning Summit Asia, which presented a great opportunity for us to gain new perspectives on approaches to technical advances in deep learning and smart AI, as well as new forms of creativity in the new technological era. It also provided us with new opportunities for business networking with other participants.

To date, the Deep Learning Summit Asia @SG 2016 is the 7th global Deep Learning Summit, and the first in Asia. It brought together industry leaders, CTOs, founders, research scientists, and entrepreneurs to explore the latest advancements and methods employed in deep learning, ranging from speech/text/image recognition to big data.

Speakers in the summit included representatives from companies and institutions such as NVIDIA, Yahoo Japan, UC Berkeley, Institute for Infocomm Research (I2R), National University of Singapore (NUS), Nanyang Technological University (NTU), Advanced Digital Sciences Center (ADSC), Google Brain, lloopp, just a few to name. They demonstrated the impact of deep learning revolution on various real applications including communications, manufacturing, healthcare and transportation.

The next Deep Learning Summit Singapore is scheduled to take place on April 27-28, 2017.

TechInnovation @SG 2016

Sep 20, 2016 | Event

Panasonic R&D Center Singapore partook in TechInnovation 2016 as one of the exhibitors to appeal to local collaboration through crowd-sourcing and crowd-pitching. TechInnovation has been a technology-industry brokerage event organized by IPI Singapore since 2012. It serves as a platform to bring together international and Singapore technology providers and seekers to explore technology and business collaboration opportunities through open innovation, and this year it successfully attracted over 2,700 participants from about 40 nations. At this, PIPM (Panasonic Intellectual Property Management Co., Ltd.), as Panasonic R&D Center Singapore’s co-exhibitor, hit a great success in attracting many enquiries and interests in their technology offer through licensing, while IPG of Panasonic R&D Center Singapore gained extensive exposure and further established a presence in the local technology market.

IP Week @ SG 2016

Aug 22, 2016 | Event

The Intellectual Property Group (IPG) of Panasonic R&D Center Singapore (PRDCSG) participated, as one of the exhibitors, in IP Ecosystem Fair, which is a part of the IP Week @ SG 2016 events organized by the Intellectual Property Office of Singapore (IPOS) from August 22nd to 24th. 2016 marked the fifth run of IP Week in Singapore and attracted over 1,000 participants from close to 30 countries, including government officials, leaders from the IP community as well as global business partners to grace the occasion, and offer their insights into the latest trends and developments in IP. While providing abundant networking opportunities among the exhibitors and participants, the fair also serves as a one-stop platform for the ‘Best of Class’ IP intermediaries and service providers to showcase their service offerings.