Expert in video compression, streaming, and processing technologies, with about 20 years of lasting impact across industry and international standardization bodies. Demonstrated ability to lead complex, cross-organizational initiatives and deliver tools that shape the streaming landscape.
June 2015 – present
Co-chair of the Video Codec Working Group in AOMedia, leading the work on the AV2 video codec.
Editor and major contributor to the AOMedia film grain synthesis 1 (AFGS1) specification.
Film grain synthesis (FGS): designed and achieved FGS adoption in AV1, led the preparation work for FGS productization at Netflix, and contributed to developing the pipeline design. The work resulted in over 30% bitrate reduction on grainy titles.
Representing Netflix as Alternate Director on the AOMedia Board of Directors.
Tech lead for the Netflix video codec team contributing to the SVT-AV1 (a joint Intel-Netflix AV1 encoder/decoder project):
University collaboration
Coordinating Netflix contributions to video technologies beyond AV1 in Alliance for Open Media (AOM).
Main author of the AV2 deblocking filter bringing close to 1% coding gain and subjective quality improvements to AV2.
Main author of the Banding prevention approach in video coding.
Proposed improvements to the AV2 film grain synthesis (FGS) algorithm that improve the synthesized grain visual quality.
Drafted the idea for generalized S-frames and RAS frames in AV2.
Delivered algorithms, encoding recipes, and software for two prominent projects in Netflix video encoding pipeline:
Active contributor to AV1 video codec
Netflix representative in video coding standardization in ITU-T (VCEG) and MPEG:
Contributor to NetVC efforts in IETF:
Feb 2008 – May 2015
Positions at Ericsson Research:
Experienced Researcher → Senior Researcher → Project Manager → Master Researcher
Led a group of Ericsson Research engineers working on subjective quality improvements for Ericsson video encoders (in collaboration with Ericsson TV, UK). The work resulted in noticeable improvements in subjective quality of HEVC and H.264 encoders on several types of video content.
Led a 3D video research project (distributed between Stockholm and Lund sites) which:
Held several leadership roles in HEVC standardization committee (JCT-VC), such as:
Author of a number of algorithms in HEVC and 3D video HEVC extensions (deblocking filter, intra-prediction, 3D reference displays SEI, reference picture list construction, non-normative encoder improvements). The HEVC deblocking filter contribution brought over 1% coding gain; later, the core design has been adopted to VVC.
Contributed algorithms to TENTM (Tandberg, Ericsson and Nokia) submission to HEVC Call for Proposals
Contributed to MPEG ISO Media Base File Format
Jan 2002 – Jan 2008
Worked on algorithms for error-resilient coding of images, video, 3D video and 3D graphics (meshes) (in particular, on multiple description coding).
Tampere University of Technology, Tampere, Finland
GPA 4.76 (0.0 to 5.0)
Thesis: "Multiple description coding of visual information". Major: Signal Processing. Minor: Multimedia
Urals State Technical University, Yekaterinburg, Russia
GPA 4.94 (summa cum laude) (2.0 to 5.0)
Thesis: "Data-link layer protocols model". Major: Computers, Multiple Computer Complexes and Networks
| Programming languages: | C, C++, Python (+Keras), Matlab, Java |
|---|---|
| Video coding standards: | AV2, AV1, VVC/H.266, HEVC/H.265, H.264, VP9, MV-HEVC, SHVC, SVC, ISO BMFF |
| Networking / protocols: | TCP/IP, UDP etc. |
| Version control: | git, SVN, CVS |
| Public speaking: | Talks |
| Expertise: | video compression, cloud-based video encoding, High Dynamic Range (HDR) video, 3D video, multi-view geometry, video processing, subjective video quality, machine learning, neural networks, CNN, error-resilience |