Research | MIT CSAIL

Time 2020-09-27 01:04:28

Web Name: Research | MIT CSAIL

WebSite: http://groups.csail.mit.edu

ID:61019

Keywords:

Research,MIT,CSAIL,

Description:

Our research centers on digital manufacturing, 3D printing and computer graphics, as well as computational photography and displays, and virtual humans and robotics. We develop new machine learning techniques and algorithms to model the transcriptional regulatory networks that control gene expression programs in living cells. We combine methods from computer science, neuroscience and cognitive science to explain and model how perception and cognition are realized in human and machine. Our group focuses on synthetic image generation, computational photography, digital manufacturing, and geometry. We use machine learning and computer vision to improve outcomes in medicine, finance, and sports We aim to develop the science of autonomy toward a future with robots and AI systems integrated into everyday life, supporting people with cognitive and physical tasks. The Genesis Group focuses on developing an account of human intelligence with specialemphasis on developing computational models of how people tell, perceive and understandstories. We are an interdisciplinary group of researchers blending approaches from human-computer interaction, social computing, databases, information management, and databases. We work at the intersection of human computer interaction and personal fabrication tools. We develop computing systems for creative expression, cultural analysis, and social change. The Interactive Robotics Group aims to imagine the future of work by designing collaborative robot teammates that enhance human capability. We conduct interdisciplinary research aimed at discovering the principles underlying the design of artificially intelligent robots. We work on a variety of topics spanning theoretical foundations, algorithms, and applications. Our projects are centered around the problems of navigation and mapping for autonomous mobile robots operating in underwater and terrestrial environments. We develop new algorithms for medical image analysis and visualization of medical imagery. We are working to elevate robots from mechanical creations controlled by low-level scripts with a considerable amount of human guidance to truly cognitive robots. We are studying how best to implement bilateral control and make it acceptable to drivers To achieve high-quality photo lighting in challenging environments, our prototype camera dynamically reconstructs a 3D scene model and directs a motor-controlled flash head at nearby walls and ceilings for soft indirect illumination. We ve developed an algorithm to anticipate visual events that may happen in the future We ve developed an object-based neural network architecture for learning predictive models of intuitive physics that extrapolates to variable object count and variable scene configurations with only spatially and temporally local computation. Our goal is to build an AI-powered personal digital nutritionist that enables users to track the food they eat simply by speaking or typing natural English phrases. We develop a computational model that explains how people make causal judgments in physical scenes by mentally simulating counterfactual outcomes and comparing those to what actually happened. By observing driver in real road situations, we learn a computer assistant that can make driving easier, safer and more enjoyable. We propose a novel aspect-augmented adversarial network for cross-aspect and cross-domain adaptation tasks. The effectiveness of our approach suggests the potential application of adversarial networks to a broader range of NLP tasks for improved representation learning, such as machine translation and language generation. Our goal is to build a system that predicts where people are looking in images. Given an image and the location of a head, our approach follows the gaze of the person and identifies the object being looked at. This project shows that we can successfully use predictive modeling to enable a database cluster to elastically expand or contract in anticipation of changes in the workload. We developed a new algorithm to generate compact sequence sets covering all k-mers using joker characters. Many modern Bayesian models involve infinitely many latent parameters. We seek to develop finite approximations which are more tractable for use in practice, and characterize their incurred error. Most popular, tractable statistical models for network data inherently assume the network is dense, although this is rarely true in practice; we propose a new modeling framework that correctly captures sparse networks. Our goal is to create an online risk-aware planner for vehicle maneuvers that can make driving safer and less stressful through a “parallel” autonomous system that assists the driver by watching for risky situations, and by helping the driver take proactive, compensating actions before they become crises. We re developing a flexible, high-performance storage architecture for database-backed applications, based on a dynamic set of queries specified by the developer which Soup automatically optimizes. We aim to understand 3D object structure from a single image. We propose an end-to-end framework which sequentially estimates 2D keypoint heatmaps and 3D object structure, by training it on both real 2D-annotated images and synthetic 3D data and by integrating a 3D-to-2D projection layer. Printable Hydraulics allows fluid-actuated robots to be automatically fabricated using 3D printers. In our work we developed a model that is able to synthesize many probable future frames with just a single image as input. Solid aims to radically change the way Web applications work today, resulting in true data ownership as well as improved privacy. Our goal is to develop an adaptive storage manager for analytical database workloads in a distributed setting. It works by partitioning datasets across a cluster and incrementally refining data partitioning as queries are run. We build tools to allow a community of people to collectively summarize large discussions online and manage knowledge embedded within these discussions. Monitoring sleep positions for a healthy rest Wireless device captures sleep data without using cameras or body sensors; could aid patients with Parkinson’s disease, epilepsy, or bedsores. Examining racial attitudes in virtual spaces through gaming Computational model developed at MIT builds off of 2019 research examining colorblind racial attitudes through a video game. Toward a machine learning model that can reason about everyday actions Researchers train a model to reach human-level performance at recognizing abstract concepts in video. Helping companies prioritize their cybersecurity investments y securely aggregating sensitive data from cyber-attacks, MIT CSAIL’s platform can quantify an organization’s level of security and suggest what to prioritize NSF awards MIT $12.5m for data science initiative w/Berkeley The National Science Foundation has awarded $12.5 million to establish a multidisciplinary institute—a collaboration between UC Berkeley and MIT—to improve our understanding of critical issues in data science, including modeling, statistical inference, computational efficiency, and societal impacts. Device for nursing homes can monitor residents activities with permission (and without video) The team s system can identify what activity a person is doing, whether that’s sleeping, reading, cooking or watching TV - and all without capturing any sensitive visual data of the activities.  Detecting and responding to incidents with images Researchers from MIT and QCRI have developed a computer vision model capable of detecting incidents in images posted on social media platforms, such as Twitter and Flickr. CSAIL to launch new initiative for machine learning applications CSAIL will be launching a new initiative for machine learning applications, titled MLA@CSAIL CSAIL co-launches virtual summit on AI in healthcare Inspired by recent events, the October 1-2 summit is being organized in collaboration with STEMM Global Scientific Community Shrinking deep learning’s carbon footprint Through innovation in software and hardware, researchers move to reduce the financial and environmental costs of modern artificial intelligence. Data systems that learn to be better Storage tool adapts to what its datasets’ users want to search. 3 Questions: John Leonard on the future of autonomous vehicles MIT Task Force on the Work of the Future examines job changes in the AV transition and how training can help workers move into careers that support mobility systems. AI system infers music from silent videos of musicians researchers describe an AI system — Foley Music — that can generate “plausible” music from silent videos of musicians playing instruments. They say it works on a variety of music performances and outperforms “several” existing systems in generating music that’s pleasant to listen to. Hundreds join MIT App Inventor’s virtual hackathon Event uses platform for non-coder app development to tackle issues such as covid-19, climate change and social justice AI systems that work w/doctors and know when to step in CSAIL s machine learning system can look at chest X-rays to diagnose pneumonia — and also knows when to defer to a radiologist. Recent advances give theoretical insight into why deep learning networks are successful. Algorithm finds hidden connections between paintings at the Met A team from MIT helped create an image retrieval system to find the closest matches of paintings from different artists and cultures. Potential COVID-19 vaccines get a boost from machine learning Future animal testing will show whether the CSAIL team’s approach could be used for human testing of a vaccine. Better simulation meshes well for design software New work on 2D and 3D meshing aims to address challenges with some of today’s state-of-the-art methods. Rewriting history to show the danger of deepfakes Can you recognize a digitally manipulated video when you see one? It’s harder than most people realize. As the technology to produce realistic “deepfakes” becomes more easily available, distinguishing fact from fiction will only get more challenging. Wu receives ACM Doctoral Dissertation Honorable Mention award ACM, the Association for Computing Machinery announced this week that MIT CSAIL PhD student ‘19 Jiajun Wu was selected for an honorable mention for his dissertation “Learning to See the Physical World.” The computational limits of deep learning Researchers warn that deep learning is reaching its computational limits. The challenge that motivates the ANA group is to foster a healthy future for the Internet. The interplay of private sector investment, public sector regulation and public interest advocacy, as well as the global diversity in drivers and aspirations, makes for an uncertain future. We devise new mathematical tools to tackle the increasing difficulty and importance of problems we pose to computers. Our vision is data-driven machine learning systems that advance the quality of healthcare, the understanding of cyber arms races and the delivery of online education. We design software for high performance computing, develop algorithms for numerical linear algebra, and research random matrix theory and its applications. This CoR brings together researchers at CSAIL working across a broad swath of application domains. Within these lie novel and challenging machine learning problems serving science, social science and computer science. Our main goal is developing a computationally based understanding of human intelligence and establishing an engineering practice based on that understanding. The MIT Center for Deployable Machine Learning (CDML) works towards creating AI systems that are robust, reliable and safe for real-world deployment. We focus on furthering the application of technology and artificial intelligence in medicine and health-care. Our group is interested in using machine learning and artificial intelligence to transform health care. This CoR aims to develop AI technology that synthesizes symbolic reasoning, probabilistic reasoning for dealing with uncertainty in the world, and statistical methods for extracting and exploiting regularities in the world, into an integrated picture of intelligence that is informed by computational insights and by cognitive science. We focus on finding novel approaches to improve the performance of modern computer systems without unduly increasing the complexity faced by application developers, compiler writers, or computer architects. Our interests span quantum complexity theory, barriers to solving P versus NP, theoretical computer science with a focus on probabilistically checkable proofs (PCP), pseudo-randomness, coding theory, and algorithms. Our lab focuses on designing algorithms to gain biological insights from advances in automated data collection and the subsequent large data sets drawn from them. Our mission is fostering the creation and development of high-performance, reliable and secure computing systems that are easy to interact with. We seek to understand the mechanistic basis of human disease, using a combination of computational and experimental techniques. Our group’s goal is to create, based on such microscopic connectivity and functional data, new mathematical models explaining how neural tissue computes. Our research centers on digital manufacturing, 3D printing and computer graphics, as well as computational photography and displays, and virtual humans and robotics. We develop new machine learning techniques and algorithms to model the transcriptional regulatory networks that control gene expression programs in living cells. We combine methods from computer science, neuroscience and cognitive science to explain and model how perception and cognition are realized in human and machine. Our group focuses on synthetic image generation, computational photography, digital manufacturing, and geometry. We develop techniques and tools that exploit automated reasoning and large amounts of computing power to tackle challenging programming problems This community is interested in understanding and affecting the interaction between computing systems and society through engineering, computer science and public policy research, education, and public engagement. We study the problem of 3D object generation. We propose a novel framework, 3D Generative Adversarial Network (3D-GAN), leveraging recent advances in volumetric convolutional networks and generative adversarial nets. Using our inkjet printer system, the goal of this project is to produce a light sensor / solar cell that is composed entirely of 3D printed materials. (This project is no longer active.) The T-1000, a prototype system of a thousand realistic processors embedded throughout an ensemble of interconnected FPGAs, seeks to demonstrate the scalability of timestamp-based cache coherence protocols on distributed shared memory systems. We develop a computational model that explains how people make causal judgments in physical scenes by mentally simulating counterfactual outcomes and comparing those to what actually happened. We design a new all-to-all broadcasts scheme with significantly less communication cost using aggregate signatures. CilkS is a new runtime system for the Cilk multithreaded programming environment which makes it easy to experiment with new algorithms, data structures, and programming linguistics. We aim to develop a systematic framework for robots to build models of the world and to use these to make effective and safe choices of actions to take in complex scenarios. Our goal is to develop an adaptive storage manager for analytical database workloads in a distributed setting. It works by partitioning datasets across a cluster and incrementally refining data partitioning as queries are run. Our goal is to understand the nature of cyber security arms races between malicious and bonafide parties. Our vision is autonomous cyber defenses that anticipate and take measures against counter attacks. Develop a method for sending persistently backlogged traffic on the Internet while adapting to link conditions dynamically. AIRvatar is a system that telemetrically collects and analyzes fine-grained data on users’ virtual identities and the process used to create them. We work on improving the algorithms for algebraic problems like matrix multiplication, and using these to design algorithms for fundamental non-algebraic problems. Alloy is a language for describing structures and a tool for exploring them. It has been used in a wide range of applications from finding holes in security mechanisms to designing telephone switching networks. Hundreds of projects have used Alloy for design analysis, for verification, for simulation, and as a backend for many other kinds of analysis and synthesis tools, and Alloy is currently being taught in courses worldwide. Amoeba is a distributed storage system that efficiently supports ad-hoc and exploratory analytics using adaptive data partitioning We are developing an algorithmic theory for brain networks, based on simple synchronized stochastic graph-based neural network models. Self-driving cars are likely to be safer, on average, than human-driven cars. But they may fail in new and catastrophic ways that a human driver could prevent. This project is designing a new architecture for a highly dependable self-driving car. Developing new methods for analyzing large quantities of player profile data, such as hundreds of thousands of profile images or chat logs. We propose efficient and effective algorithms to perform approximate string joins with abbreviations in database systems. There is a family of approximation algorithms for computing the diameter of an undirected graph that give a time/accuracy trade-off and our goal is to extend these results to directed graphs. The Arabic language is spoken by over one billion people around the world. Arabic presents a variety of challenges for speech and language processing technologies. In our group, we have several research topics examining Arabic, including dialect identification, speech recognition, machine translation, and language processing. By observing driver in real road situations, we learn a computer assistant that can make driving easier, safer and more enjoyable. Our goal is to develop a socially intelligent team coacher agent that helps humans communicate, strategize, and work together efficiently. Monitoring sleep positions for a healthy rest Wireless device captures sleep data without using cameras or body sensors; could aid patients with Parkinson’s disease, epilepsy, or bedsores. Examining racial attitudes in virtual spaces through gaming Computational model developed at MIT builds off of 2019 research examining colorblind racial attitudes through a video game. Toward a machine learning model that can reason about everyday actions Researchers train a model to reach human-level performance at recognizing abstract concepts in video. Helping companies prioritize their cybersecurity investments y securely aggregating sensitive data from cyber-attacks, MIT CSAIL’s platform can quantify an organization’s level of security and suggest what to prioritize NSF awards MIT $12.5m for data science initiative w/Berkeley The National Science Foundation has awarded $12.5 million to establish a multidisciplinary institute—a collaboration between UC Berkeley and MIT—to improve our understanding of critical issues in data science, including modeling, statistical inference, computational efficiency, and societal impacts. Device for nursing homes can monitor residents activities with permission (and without video) The team s system can identify what activity a person is doing, whether that’s sleeping, reading, cooking or watching TV - and all without capturing any sensitive visual data of the activities.  Detecting and responding to incidents with images Researchers from MIT and QCRI have developed a computer vision model capable of detecting incidents in images posted on social media platforms, such as Twitter and Flickr. CSAIL to launch new initiative for machine learning applications CSAIL will be launching a new initiative for machine learning applications, titled MLA@CSAIL CSAIL co-launches virtual summit on AI in healthcare Inspired by recent events, the October 1-2 summit is being organized in collaboration with STEMM Global Scientific Community Shrinking deep learning’s carbon footprint Through innovation in software and hardware, researchers move to reduce the financial and environmental costs of modern artificial intelligence. Data systems that learn to be better Storage tool adapts to what its datasets’ users want to search. 3 Questions: John Leonard on the future of autonomous vehicles MIT Task Force on the Work of the Future examines job changes in the AV transition and how training can help workers move into careers that support mobility systems. AI system infers music from silent videos of musicians researchers describe an AI system — Foley Music — that can generate “plausible” music from silent videos of musicians playing instruments. They say it works on a variety of music performances and outperforms “several” existing systems in generating music that’s pleasant to listen to. Hundreds join MIT App Inventor’s virtual hackathon Event uses platform for non-coder app development to tackle issues such as covid-19, climate change and social justice AI systems that work w/doctors and know when to step in CSAIL s machine learning system can look at chest X-rays to diagnose pneumonia — and also knows when to defer to a radiologist. Recent advances give theoretical insight into why deep learning networks are successful. Algorithm finds hidden connections between paintings at the Met A team from MIT helped create an image retrieval system to find the closest matches of paintings from different artists and cultures. Potential COVID-19 vaccines get a boost from machine learning Future animal testing will show whether the CSAIL team’s approach could be used for human testing of a vaccine. Better simulation meshes well for design software New work on 2D and 3D meshing aims to address challenges with some of today’s state-of-the-art methods. Rewriting history to show the danger of deepfakes Can you recognize a digitally manipulated video when you see one? It’s harder than most people realize. As the technology to produce realistic “deepfakes” becomes more easily available, distinguishing fact from fiction will only get more challenging. Wu receives ACM Doctoral Dissertation Honorable Mention award ACM, the Association for Computing Machinery announced this week that MIT CSAIL PhD student ‘19 Jiajun Wu was selected for an honorable mention for his dissertation “Learning to See the Physical World.” The computational limits of deep learning Researchers warn that deep learning is reaching its computational limits.

TAGS:Research MIT CSAIL 

<<< Thank you for your visit >>>

Hot Websites