Top 10 Intel Labs Blogs for 2020

Highlights

  • Find out about Intel Labs’ Top 10 blogs for 2020.

  • From research papers to faculty awards programs to new collaborative institutes, Intel Labs stories captured readers’ attention in 2020.

author-image

投稿者:

As data becomes more deeply embedded in our lives with every technology leap — from AI to 5G to the intelligent edge — Intel Labs’ role is to find new ways to unleash its full potential, according to Rich Uhlig, Intel senior fellow, vice president and director of Intel Labs.

As data becomes more deeply embedded in daily life with every technology leap — from artificial intelligence (AI) to 5G to the intelligent edge — Intel Labs’ role is to find new ways to unleash its full potential. In 2020, Intel Labs focused on the themes of breaking down data barriers, moving beyond today's AI, using technology for good, and building trust through confidential computing.

From deep learning for malware detection to advances in robot surgery to the newly established Cognitive Computing Research group, Intel Labs blogs covered these themes in blogs published in 2020. From research papers to faculty awards programs to new collaborative institutes, Intel Labs stories on innovative technology captured readers’ attention.

Here are the Top 10 Intel Labs blogs for 2020 (in order of publication):

1. STAMINA: Deep Learning for Malware Detection
Researchers from Intel Labs and Microsoft Threat Protection Intelligence Team joined forces to study the use of deep learning for malware threat detection. The joint team studied how to apply deep transfer learning from computer vision to static malware classification.

2. Cognitive Computing Research: From Deep Learning to Higher Machine Intelligence
Intel Labs' newly established Cognitive Computing Research will drive innovation in machine intelligence and cognition. Combining deep learning with knowledge graphs and symbolic reasoning, the research will focus on a new class of artificial intelligence solutions to improve efficiency, explainability, extensibility, and reasoning capabilities of AI systems.

3. Researchers Use Deep Learning to Train Autonomous Acrobatic Drones
Researchers safely trained acrobatic controllers in simulation and deployed them with no fine-tuning on physical quadrotor drones using zero-shot transfer. A vision-based drone with only onboard sensing and computation can now autonomously perform agile maneuvers with accelerations of up to 3g.

4. AI Medical Robot Learns How to Suture by Imitating Videos
Imitation learning provides a promising approach to teach robotic surgical skills using expert video demonstrations. On average, the robot had 85.5% segmentation accuracy, suggesting performance improvement over several state-of-the-art baselines, while kinematic pose imitation gave 0.94 centimeter error in position per observation on the test set.

5. How Rolling Loaded Dice Will Change the Future of AI
Researchers at the MIT Probabilistic Computing Project, funded by the Intel Probabilistic Computing Center, developed a computer algorithm to simulate the roll of “loaded” dice to produce random numbers. With its low memory requirements and fast speed, the algorithm may improve the accuracy and efficiency of artificial intelligence applications, including decision support.

6. Multi-access Traffic Management at the Edge
As the mobile industry moves toward 5G, performance requirements from video streaming and other applications are increasing. Intel Lab researchers are taking on the multi-access challenge to manage data traffic across all available access networks and meet diverse application requirements in rate, latency, and reliability.

7. Intel’s 2020 Rising Star Faculty Award Program Recognizes 10 Leading Researchers
Intel’s 2020 Rising Star Faculty Award program recognizes 10 leading university researchers likely to build tomorrow’s disruptive computing technology. The program fosters long-term collaborative relationships with senior technical leaders at Intel.

8. Compute Near-Memory Reduces Data Transfer Energy and Increases Throughput in Neural Networks
Intel Labs has built an AI circuit that can alleviate memory bottlenecks by inserting distributed computational units into the memory array, creating compute near-memory systems. With built-in scalability and modularity, the AI accelerator circuit can address a range of needs from low-power IoT and edge devices all the way to data centers and servers.

9. The Three Pillars of Machine Programming Provide Core Concepts for Research Advances
In the future, Intel Labs researchers believe that computers will become programmers, where they will invent the algorithms and data structures necessary to realize a programmer’s intention through an emerging technology known as machine programming (MP). The field of MP is driven by the three pillars — intention, invention, and adaptation — which provide the conceptual framework for Intel Lab’s numerous MP research advances.

10. Newly Launched Private AI Collaborative Research Institute Funds First 9 Research Projects
Intel, in collaboration with Avast and Borsetta, launched the Private AI Collaborative Research Institute to advance and develop technologies in privacy and trust for decentralized AI. The Private AI Collaborative Research Institute selected the first nine institute-supported research projects, distributed among eight universities worldwide.