About Us

The Graphics & Media Lab Video Group is part of the Computer Science Department of Lomonosov Moscow State University. The Graphics Group began at the end of the 1980s, and the Graphics & Media Lab was officially founded in 1998. The main research avenues of the lab include areas of computer graphics, computer vision and media processing (audio, image, and video). A number of patents have been acquired based on the lab’s research, and other results have been presented in various publications. The main research avenues of the Graphics & Media Lab Video Group are video processing, video quality analysis, and video compression (codecs testing and tuning, quality metrics research, and codecs development).

Achievements

#1 video codecs analysis

The companies send us the codecs, and we find out, which one is better.

Here is a list of companies we collaborated with:

Our stats by 2019:

• 210 analysed video-codecs
• 22 published comparisons reports
• 500 thousand downloads of reports

Some of our reports are publically available.

#1 video quality estimation tool

We have VQMT for monoscopic and VQMT3D for stereoscopic video. We can predict in advance whether you will have a headache after watching the 3D movie. The tools have more quality assessment metrics than any concurrent ones declared. Take a look at our assessment reports on the projects’ pages. Our stats by 2019:

• 22 monoscopic video quality metrics
• 13 stereoscopic video quality metrics
• 150 estimated feature-length stereoscopic movies
• 9 reports on stereoscopic movies estimation

#1 video matting benchmark

The VideoMatting project is the first public objective benchmark for video-matting methods. Our work ranks existing methods and aids developers of new methods in improving their results. Previous cases of estimation:

• Comparison of video matting methods #1
  • 13 competitors
  • 30 test-cases
  • 543 participants
• Comparison of video matting methods #2
  • 14 competitors
  • 30 test-cases
  • 434 participants

#1 subjective quality estimation

The Subjectify project provides a crowd-sourced algorithms evaluation platform. It is designed for the subjective comparison of images, video, and sound processing methods. Previous cases of estimation:

• Comparison of video completion methods
  • 6 competitors
  • 7 test-cases
  • 341 participant
• Comparison of saliency-aware encoder
  • 2 competitors
  • 31 test-case
  • 98 participants
• “--tune ssim” comparison
  • 4 competitors (× 3 bitrates)
  • 4 test-cases
  • 219 participants
• Video codec comparison
  • 7 competitors (× 3 bitrates)
  • 4 test-cases
  • 325 participants (11530 answers)
• Netflix study replication
  • 6-10 competitors (for various test-cases)
  • 7 test-cases
  • 375 participants

Other

• We make algorithms to correct and improve videos. Our filters have 3 million downloads by 2019.
• The EVT project increases the video’s quality and decreases the file size, by providing the optimal parameters for any codec. Yours too.
• The Compression project is the biggest resource on video compression.
• We have developed the exclusive tools for Intel, Samsung, Huawei, RealNetworks, and other companies, adapting our algorithms for specific video streams, applications and hardware like TV-sets, graphics cards, etc.

Contacts

We’re located in Moscow State University, Moscow, Russia. We’re open to cooperation. Feel free to contact us: dmitry@graphics.cs.msu.ru.