Kaushik Balakrishnan

AI Researcher

San Francisco, California, United States15 yrs 4 mos experience

Most Likely To SwitchAI Enabled

Key Highlights

Over 10 years of experience in machine learning and AI.
Expert in high-performance computing and numerical simulations.
Proven track record in developing advanced algorithms for autonomous systems.

Stackforce AI infers this person is a Machine Learning and AI expert with a strong focus on high-performance computing in the automotive and aerospace industries.

Contact

cowshik25@rediffmail.com LinkedIn

Skills

Core Skills

Machine LearningDeep LearningComputer VisionArtificial IntelligenceApplied MathematicsHigh Performance ComputingComputational Science

Other Skills

Anomaly DetectionComputational Fluid DynamicsConvolutional Neural NetworksConvolutional-Transformer architecturesData Analysis AlgorithmsDriver State EstimationGPUGenerative Adversarial Imitation LearningGenerative Adversarial NetworksHuman Pose EstimationMasked AutoencodersNatural Language ProcessingNumerical AnalysisObject DetectionParallel Computing

About

Research Scientist with 10+ years experience in massively-parallel numerical simulations in the area of Machine Learning, Deep Learning, Artificial Intelligence, Reinforcement Learning, High Performance Computing (HPC), Monte Carlo methods, etc. Expertise in the following areas: 1. Computer programming: C, C++, Python 2. Packages: TensorFlow, PyTorch, Caffe, OpenCV, R, MATLAB, Octave 3. High-performance computing: MPI, GPU (CUDA) 4. Machine Learning: Deep Neural Networks, Convolutional Neural Networks (CNN), Generative Adversarial Networks (GAN), Variational AutoEncoders (VAE), Principal Component Analysis (PCA), Naive Bayes, Expectation Maximization, Gibbs sampling GPU (CUDA) computing: Hydrogen-Oxygen flame simulations using shared memory; Reaction-Diffusion simulations of the Schlogl and Gray-Scott models; real-time visualization using OpenGL; Reduced Basis Modeling (RBM) using POD-DEIM GitHub: https://github.com/kaushikb258

Experience

Samsung display

Senior Staff Machine Learning Engineer

Aug 2022 – Present · 3 yrs 7 mos · San Jose, California, United States · On-site

Patent relevance classification using Natural Language Processing (NLP): RoBERTa + Sentence-BERT
Image retrieval using Self-Supervised Learning: Vision Transformers + Masked Autoencoders
Anomaly detection using multi-modal data

Natural Language ProcessingSelf-Supervised LearningAnomaly DetectionMachine LearningDeep Learning

Ford motor company

Machine Learning Research Scientist

May 2019 – Jul 2022 · 3 yrs 2 mos · Palo Alto, CA

Human pose estimation using hybrid Convolutional-Transformer architectures
Self-Supervised Learning for Computer Vision
Driver State Estimation using Computer Vision
Exploration of Reinforcement Learning algorithms for indoor robot navigation and highway driving
Object Detection and Semantic Segmentation for autonomous driving applications

Human Pose EstimationSelf-Supervised LearningComputer VisionReinforcement LearningMachine Learning

Bmw group

Machine Learning Engineer - Reinforcement Learning/Artificial Intelligence

May 2018 – May 2019 · 1 yr · Mountain View, CA

Exploration of Reinforcement Learning algorithms for autonomous driving policy applications
Generative Adversarial Imitation Learning (GAIL)
Robot Operating System (ROS)
Computer Vision

Reinforcement LearningGenerative Adversarial Imitation LearningComputer VisionArtificial IntelligenceMachine Learning

Ford motor company

Autonomous Vehicles Research Scientist

Mar 2017 – Apr 2018 · 1 yr 1 mo · Palo Alto, CA

Deep learning for autonomous vehicles: Deep Neural Networks (DNN), Convolutional Neural Networks (CNN), Generative Adversarial Networks (GAN), Variational AutoEncoders (VAE), Long Short Term Memory (LSTM), Deep Reinforcement Learning
Semantic segmentation using Fully Convolutional Networks (FCN) and Generative Adversarial Networks (GAN) for autonomous vehicle applications; Autonomous Driving in a car simulator using Deep Reinforcement Learning; Automatic parking using Deep Reinforcement Learning; Localization of autonomous vehicles on a prior map using Lidar data; Object detection using YOLO and SSD for autonomous driving applications
Explored different GAN algorithms such as pix2pixGAN, Cycle-GAN, Wasserstein-GAN, LSGAN, etc.

Deep LearningGenerative Adversarial NetworksSemantic SegmentationObject DetectionMachine Learning

Hypercomp, inc.

Research Scientist

Jul 2015 – Feb 2017 · 1 yr 7 mos · Westlake Village, CA

Algorithms for efficient data analysis for engineering applications
Reduced Order Modeling of large-scale time variant data applied to pressure oscillations in rockets
Proper Orthogonal Decomposition, Singular Value Decomposition, Discrete Empirical Interpolation Methods, Greedy Algorithms, etc.

Data Analysis AlgorithmsReduced Order ModelingProper Orthogonal DecompositionApplied Mathematics

Nasa jet propulsion laboratory

Computational Scientist

Feb 2014 – Jun 2015 · 1 yr 4 mos · Pasadena, CA

Propulsion, Thermal and Materials Engineering Division
High Performance Computing of fluid-flow problems
High fidelity two-phase numerical simulation of a rocket landing on the surface of the Moon and Mars
Computational Fluid Dynamics (CFD) investigation of supersonic rocket exhaust jets

High Performance ComputingComputational Fluid Dynamics

Lawrence berkeley national laboratory

Postdoctoral Fellow, Computational Research Division

Jul 2010 – Jan 2014 · 3 yrs 6 mos · Berkeley, CA

Computational Fluid Dynamics (CFD) analysis of explosions using high-fidelity simulations
Parallel computing using MPI and OpenMP on supercomputers
Research on stochastic methods for fluid dynamic applications
Development of robust numerical schemes for multiphase flows and fluctuating hydrodynamics

Computational Fluid DynamicsParallel ComputingComputational Science