Week Ending 11.29.2020
RESEARCH WATCH: 11.29.2020
Over the past week, 853 new papers were published in "Computer Science".
The paper discussed most in the news over the past week was "How Did That Get In My Phone? Unwanted App Distribution on Android Devices" by Platon Kotzias et al (Oct 2020), which was referenced 20 times, including in the article Baidu's Android Apps Caught Collecting and Leaking Sensitive User Data in The Hacker News. The paper got social media traction with 5 shares. A user, @vir2alexport, tweeted "A large part of monetization libraries are vulnerable and directly expose users’ information in «Man In The Middle» attacks. "between 10% and 24% of users devices encounter at least one unwanted app"".
Leading researcher Yoshua Bengio (Université de Montréal) published "RetroGNN: Approximating Retrosynthesis by Graph Neural Networks for De Novo Drug Design".
The paper shared the most on social media this week is by a team at Salesforce: "CoMatch: Semi-supervised Learning with Contrastive Graph Regularization" by Junnan Li et al (Nov 2020) with 123 shares. @ChrSzegedy (Christian Szegedy) tweeted "Interesting paper: 94% on cifar-10 with 80 labeled examples (8/class)".
This week was active for "Computer Science - Artificial Intelligence", with 122 new papers.
The paper discussed most in the news over the past week was by a team at Rutgers University: "Deep learning for video game genre classification" by Yuhang Jiang et al (Nov 2020), which was referenced 5 times, including in the article Researchers taught AI how to judge a video game by its cover in The Next Web. The paper got social media traction with 13 shares. A Twitter user, @ml_india_, said "#AINews: Researchers have combined cutting-edge image recognition and NLP to create an AI system for video game genre classification based on its cover! 🕹 Details: Paper: #ArtificialIntelligence #MachineLearning #DataScience".
Leading researcher Luc Van Gool (Computer Vision Laboratory) came out with "Learning from Simulation, Racing in Reality".
The paper shared the most on social media this week is by a team at Chinese Academy of Sciences: "Towards Playing Full MOBA Games with Deep Reinforcement Learning" by Deheng Ye et al (Nov 2020) with 82 shares. @masafumi (masafumi) tweeted "リプライで情報もらったけど強化学習用の環境強いな、さすがTencent。 "Our RL infrastructure runs over a physical computing cluster, containing 250,000 CPU cores and 2,000 Nvidia V100 GPU cards. "".
This week was active for "Computer Science - Computer Vision and Pattern Recognition", with 260 new papers.
The paper discussed most in the news over the past week was by a team at Rutgers University: "Deep learning for video game genre classification" by Yuhang Jiang et al (Nov 2020)
Leading researcher Trevor Darrell (UC Berkeley) published "Temporal Action Detection with Multi-level Supervision" The investigators introduce the Semi - supervised Action Detection (SSAD) task with a mixture of labeled and unlabeled data and analyze different types of errors in the proposed SSAD baselines which are directly adapted from the semi - supervised classification task.
The paper shared the most on social media this week is by a team at Salesforce: "CoMatch: Semi-supervised Learning with Contrastive Graph Regularization" by Junnan Li et al (Nov 2020)
This week was active for "Computer Science - Computers and Society", with 31 new papers.
The paper discussed most in the news over the past week was "Understanding bias in facial recognition technologies" by David Leslie (Oct 2020), which was referenced 1 time, including in the article Report says lack of diversity in face biometrics datasets extends to expression, emotion in BiometricUpdate.com. The paper got social media traction with 5 shares.
Over the past week, 20 new papers were published in "Computer Science - Human-Computer Interaction".
The paper discussed most in the news over the past week was by a team at Carnegie Mellon University: "Batteries, camera, action! Learning a semantic control space for expressive robot cinematography" by Rogerio Bonatti et al (Nov 2020), which was referenced 2 times, including in the article AI that directs drones to film ‘exciting’ shots could lower video production costs in Venturebeat. The paper got social media traction with 12 shares. The investigators develop a data - driven framework that enables editing of these complex camera positioning parameters in a semantic space .
This week was very active for "Computer Science - Learning", with 303 new papers.
The paper discussed most in the news over the past week was by a team at Google: "Underspecification Presents Challenges for Credibility in Modern Machine Learning" by Alexander D'Amour et al (Nov 2020), which was referenced 6 times, including in the article The Way We Train AI Is Fundamentally Flawed in Zero Hedge. The paper author, Alex D’Amour, was quoted saying "We are asking more of machine-learning models than we are able to guarantee with our current approach". The paper got social media traction with 1430 shares. A user, @popular_ML, tweeted "The most popular ArXiv tweet in the last 24h", while @julius_adebayo posted "This week's new must read 30 pager. Domain shift and spurious training signals are major open problems in ML".
Leading researcher Yoshua Bengio (Université de Montréal) came out with "RetroGNN: Approximating Retrosynthesis by Graph Neural Networks for De Novo Drug Design".
The paper shared the most on social media this week is by a team at Salesforce: "CoMatch: Semi-supervised Learning with Contrastive Graph Regularization" by Junnan Li et al (Nov 2020)
Over the past week, five new papers were published in "Computer Science - Multiagent Systems".
Over the past week, 17 new papers were published in "Computer Science - Neural and Evolutionary Computing".
This week was active for "Computer Science - Robotics", with 54 new papers.
The paper discussed most in the news over the past week was by a team at Hong Kong University of Science and Technology: "Origami-based Shape Morphing Fingertip to Enhance Grasping Stability and Dexterity" by Zicheng Kan et al (Oct 2020), which was referenced 4 times, including in the article Unreal robot fingertip looks like origami and morphs like a Transformer in Syfy Wire. The paper author, Yazhan Zhang (Hong Kong University of Science and Technology), was quoted saying "A Ph.D. student in our group, Mr. Song Haoran, also previously published a paper on contact surface clustering , showing three typical contact primitives for the representations of major local geometries".
Leading researcher Dhruv Batra (Georgia Institute of Technology) published "Learning Navigation Skills for Legged Robots with Learned Robot Embeddings".