Research

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.

We work on understanding and improving vision algorithms used for self driving cars, by introducing synthetic data-sets to the loop.

Developing state-of-the-art tools that process 3D surfaces and volumes

Our goal is to understand what drivers can perceive when they are not attending to the road ahead and the consequences of their perceptions for safety.

Gerrymandering is a direct threat to our democracy, undermining founding principles like equal protection under the law and eroding public confidence in elections.

Predicting the number of clock cycles a processor takes to execute a block of assembly instructions in steady-state (the throughput) is important for both compiler designers and performance engineers.

However, building an analytical model to do so is especially complicated in modern x86-64 Complex Instruction Set Computer (CISC) machines with sophisticated processor microarchitectures in that it is tedious, error-prone, and must be performed from scratch for each processor generation.

Ithemal is the first tool that learns to predict the throughput of a set of instructions. It does so more accurately than state-of-the-art hand-written tools currently used in compiler backends and static machine code analyzers. In particular, Ithemal has less than half the error of state-of-the-art analytical models (LLVM's llvm-mca and Intel's IACA).

Our goal is to create a theoretical framework and effective machine learning algorithms for robust, reliable control of autonomous vehicles. Key threads include developing metrics of confidence; and designing deep learning algorithms for parallel autonomy.

In this project, we aim to develop a framework that can ensure and certify the safety of an autonomous vehicle. By leveraging research from the area of formal verification, this framework aims to assess the safety, i.e., free of collisions, of a broad class of autonomous car controllers/planners for a given traffic model.

We want to help to make driving more comfortable, environmentally friendly and above all safer!

Developing state-of-the-art deep learning algorithms for analyzing and modeling 3D geometry

A polyhedral compiler for expressing image processing, DNN, and linear/tensor algebra applications