Arslan Ali
Senior Ai Applied Research Scientist (Genai - Cosmos) @NVIDIA
Signup · Get unlimited contacts
WORK HISTORY
Senior Ai Applied Research Scientist (Genai - Cosmos) @NVIDIA
DE
Lead Generative AI from research to production, including Text2image, image2image, in-painting, image segmentation, depth2image models. • Designed and developed highly scalable end-to-end AI workflow for the highest quality restoration of old and deteriorated footage. Leading the R&D work for AI-based super-resolution, de-interlacing, and pre-and post-processing algorithms. • Led the R&D work for the replacement of AV1 encoder internal state predictor with CNNs for neighbor block prediction.• Hyper-scale XR: Designed and developed Equivariant Neural Renderers, Auto encoders and Denoisers on distributed clusters SMP and GPU-based modules for VR head-mounted displays (HMD) for virtual telepresence, conferencing, and virtual entertainment. • Wrote NVIDIA blog series on Efficient deployment of diverse AI model categories (classification, detection, segmentation, generative & stateful models) from public model zoo using NVIDIA Triton inference server along with efficient horizontal and vertical scaling. https://bit.ly/3iK2Ifw • Led a team to develop a cloud-native GStreamer-based stack that applies AI-processing on live video streams. Presented GTC talk, which partially showcases an earlier version of the aforementioned project. https://bit.ly/3ZEFM1T • Expert of low-level profiling, production level performance analysis and optimization of AI models and end-to-end pipelines. Increased performance of deep-neural nets for major studio software titles by 2-6x. • Designed and developed an automated profiling tool agnostic to architectures, applications, and custom ISV builds for an easier comparison of underlying hardware and estimation of application optimization potentials. • Helped NVIDIA partners to adapt:o Triton Inference Server: that simplifies and accelerates the AI model deployment supporting all major deep learning backends.DeepStream SDK: is a streaming analytical toolkit for the rapid development and deployment of Vision AI applications.
EDUCATION
Center for Advanced Studies in Engineering
BS electrical engineering, Computer
University of Stuttgart
Master’s Degree, Information Technology specialisation in Embedded Systems
Politecnico di Torino
Doctor of Philosophy - PhD, Electrical, Electronics and Communications Engineering
ABOUT ARSLAN ALI
AI Research Scientist and Engineer with eight years of industry and academia R&D experience. Specializes in generative, classification, detection, and segmentation models with demonstrated computer vision, image/video processing, and biometrics proficiency. Extensive knowledge in computer vision, image and video processing, biometrics, signal processing, cryptography, compressive sensing, robotics, real-time operating systems, horizontal and vertical scaling, Docker, Kubernetes, cloud computing and deployment, profiling, multicore multiprocessor systems, and GPUs. Notably, led groundbreaking projects at multinational organizations, leading to several advancements in AI. Holds numerous patents and has an impressive record of publications in prestigious conferences and journals. Hold notable academic achievements, including a Gold medal and Summa Cum Laude.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.