Aug 2020 – Jun 2023 · 2 yrs 10 mos · San Diego, California, United States · On-site

• Platform Architecture in HW Tech :
• Image/Video and Multimedia-Processing - pixel pipe design
• Defining SoC architectures( HW or SW and intersection thereof) for display rendering use-cases
• Display technologies and picture-quality
• Streamlined Image/Video-Processing /Computer-Vision algorithm development for Apple products
• Algorithm and system optimizations, trade-offs

Mar 2018 – Aug 2020 · 2 yrs 5 mos

• Core member of HDR10+ and Imaging R&D team
• Active R&D, contributions to standardization
• HDR10+ : https://hdr10plus.org/
• SMPTE ST2094-40
• Author of HDR10+ reference code for adopter SoC companies
• Industry demos, design certification tests, contribution to tech spec
• Lead Architect of HDR Video platform solution for HW/SoC chipset
• Design cost-efficient & optimized HW/SW architecture SDR-HDR media processing
• HDR10+, HDR10 PQ/HLG, ST2094-40, Metadata parsing
• Dynamic Tone Mapping
• Inverse Tone Mapping
• HDR and wide color gamut perceptual enhancements
• Color processing, convesion & Gamut-Mapping
• Lead developer in commercialization of HDR rendering systems and HDTV chipsets
• Algorithms for HDR10+ source workflow tools
• HDR10+ sink side color-volume mapping
• Deep-Learning techniques for next generation AI based pipelines.
• Perceptual image quality improvements
• Hybrid pixel-processing algorithms (part conventional, part DL based)
• Propose/invent project ideas, roadmaps and build to production
• Computer Vision applications for next generation 8K Samsung DTVs
• AI & Deep-Learning: Prepare proposals & POCs

Jul 2014 – Mar 2018 · 3 yrs 8 mos

• Algorithm PoC to commercialization: Apply state-of-the art Machine-Learning, Image-Processing techniques
• Lead developer :
• Banding & Quantization artifact detection and removal
• Video De-Noising, compression artifact removal
• ML based feature extraction & classification for various Video Enhancement blocks
• SDR to HDR : Inverse Tone Mapping
• Image bit-depth expansion: 8 to 10/12 bit precision
• HDR Perceptual Peak, Black Enhancement
• HDR Detail Enhancement
• Samsung's Dynamic TM curve optimization (1st gen version adopted in 2017 models)
• Lead efforts in commercialization of HDR features for 2017~2020 & future HDTV chipsets
• Actively propose new project ideas
• Computer Vision applications for next generation 8K Samsung DTVs
• AI & Deep-Learning: Prepare proposals & POCs

Dec 2009 – Jan 2013 · 3 yrs 1 mo · Irvine, California

• Research and development of advanced algorithms in Video-Processing for HDTV and visual displays. Image/Video-Processing, Computer-Vision algorithms for HDTV, SmartTV, 3DTV products.
• Super-Resolution / Machine-Learning based Image up-scaling
• Image and Object Segmentation for 2D-to-3D depth estimation
• Object-Tracking for Image tool workflows
• Video de-noising : 3D video filtering and flicker reduction
• De-noising and color artifact reduction in low quality internet videos
• Color gamut-mapping for HDTV display
• Development cycles
• Algorithm development and C++ prototyping.
• Optimizations - Hardware efficient implementations (for SoC . ASIC based solutions)
• GPU / CUDA implementation for performance efficiency.

Jan 2013 – Jun 2014 · 1 yr 5 mos · Greater Seattle Area

• Part of Computer-Vision and Machine-Learning team innovating and developing solutions ground-up for improving processes, security and productivity using intelligent Imaging and Vision technologies across Fulfilment Centers.

Jul 2009 – Sep 2009 · 2 mos · Portland, Oregon Area

• Image Processing related to LCD/LED HDTV
• Algorithms for automatic Mura (LCD defect) detection.
• R&D on multi-primary color reproduction in LED backlit LCD HDTV.

May 2008 – Aug 2008 · 3 mos · Greater Minneapolis-St. Paul Area

• Research on Low Frequency & Periodic Noise Cancellation for Biomedical/Hearning Aid Applications

Aug 2007 – Dec 2009 · 2 yrs 4 mos · Phoenix, Arizona Area

• Research Assistant at Visual Representation & Processing Group
• Compressive Sensing (CS) method for Imaging, Video Coding
• Demosaicing for Compressive Sensing (CS) image acquisition
• Single Image Super-Resolution : Sparse representation and dictionary based learning methods.
• Face-Recognition in low dimension : feature extraction and sparse learning based classification competitive with state-of-the art.
• Fast GPU / CUDA implementation by parallel computing for sparse learning based imaging applications
• Developed novel algorithms for joint coding of RGB color images in mage acquisition frameworks.

Jul 2004 – Aug 2007 · 3 yrs 1 mo · Bengaluru Area, India

• Technical Center India
• Design and develop embedded system software for multimedia applications Satellite Digital Audio Receivers (SDARs), Audio/Video playback systems for leading automotive manufacturers.
• Gained experience in full software-development cycle from requirement analysis, design, implementation through testing on various embedded - RISC/DSP platforms.
• Developed strong understanding of architecture for real-time embedded systems and multimedia software/system level design problems, implementation and testing issues.
• Worked seamlessly with full product teams across the globe (U.S.A, Singapore, Mexico and Germany) – collaborated with electrical design, algorithms, systems and manufacturing teams and gained expertise on broader aspects of "Engineering Multimedia Products”.

Jan 2004 – May 2004 · 4 mos · Bengaluru Area, India

• Technical Center India, Bangalore.
• Adaptive Acoustic Echo Canceller (AEC): An Audio/Speech application. (MATLAB / TI DSP 5509)
• Designed and implemented an adaptive AEC that models the car acoustic environment in real-time (TI 5509 DSP platform). The purpose was to track & model the echo generated from acoustic audio output and retain desired signals like speech. Such systems were front end pre-processing units for Automatic Speech Recognition Engines.