Advanced Media Lab

Elevating the media experience

Ofinno’s Advanced Media Lab is a top contributor to MPEG and JVET — the standards bodies defining the future of video compression and immersive media. Our researchers are actively shaping the next-generation video coding standard beyond VVC, as well as standards for 3D point cloud compression, 3D mesh compression, and 3D Gaussian splatting.

Compression built for what comes next

Video accounts for the vast majority of internet traffic, and the demand for more efficient compression grows with every jump in resolution, every new immersive format, and every device added to the network. But the future of media compression goes well beyond traditional 2D video. Ofinno’s researchers are contributing across JVET and multiple MPEG working groups, advancing the coding tools that will define the next generation of video compression while simultaneously developing standards for entirely new media formats — 3D point clouds, dynamic meshes, and Gaussian splatting — that will power the immersive experiences of tomorrow.

Where we contribute

Next-Generation Video Coding (JVET)

Intra prediction & template matching, spatial geometric partition modes, intra block copy & block vector prediction, quantization & residual coding

Point Cloud Compression (G-PCC)

TriSoup geometry coding, vertex position entropy encoding, RAHT attribute coding, ACU-level quantization, inter-frame motion-compensated prediction

3D Mesh Compression (V-Mesh)

Wavelet coefficient packing, displacement quantization & coding, weighted coefficient prediction, dynamic mesh attribute processing

3D Gaussian Splatting (3DGS)

Splat attribute coding, dequantization offset adjustment, volume-based QP adaptation, frame/tile/slice signaling structure

Standards Readouts

Read our latest posts about standards development in multimedia and streaming

More Readouts
Our Team

Meet Ofinno’s multimedia and streaming technology experts

More Team Members


Our Team

Meet Ofinno’s cellular network technology experts

Meet our team


We are defining what comes next in media compression

We are defining what comes next in media compression

Building the Video Codec Beyond VVC

VVC (H.266) is the current state of the art in video compression, but its successor is already taking shape within JVET. Ofinno’s researchers are among the most active contributors to this effort. Our work focuses on making video compression smarter and more efficient — advancing template-based intra prediction that can derive optimal coding modes without explicit signaling, spatial geometric partition modes that enable more flexible block partitioning, and block vector intelligence that uses reference content not just for copy but as a signal for mode derivation and reference sample guidance. These innovations, and others, target meaningful improvements in compression efficiency at every level of the coding pipeline.

Compressing the 3D World — Point Clouds and Meshes

As immersive media moves from concept to reality, the standards for compressing 3D content are being written now. Ofinno’s researchers are contributing to both G-PCC (geometry-based point cloud compression) and V-Mesh (3D mesh compression) under MPEG WG7. In G-PCC, our work spans TriSoup geometry coding with innovations in vertex position entropy encoding and occupancy tree coding, RAHT-based attribute coding with adaptive quantization, and inter-frame prediction using motion-compensated attribute projection onto reconstructed geometry. In mesh compression, our contributions cover wavelet coefficient packing for vertex displacements, displacement quantization, and weighted coefficient prediction for dynamic 3D mesh content. Together, these contributions address the full compression pipeline — both intra-frame and inter-frame — for the dominant 3D content representations.

Pioneering Gaussian Splatting Compression

3D Gaussian splatting is one of the most rapidly emerging techniques for representing and rendering photorealistic 3D scenes. Ofinno’s researchers are contributing to MPEG’s exploration of Gaussian splatting compression, developing techniques for splat attribute coding with QP adaptation, dequantization offset adjustment for attribute coefficients, and volume-based quantization parameter adjustment. Ofinno is positioned at the forefront of this technology that could fundamentally change how immersive 3D content is stored and delivered.

Related Content

Read more about our work in this area

All insights

Let’s collaborate

Interested in working with Ofinno to develop new innovations or acquire patents to further your goals? Let’s talk. Send us a message and we’ll connect soon.

"*" indicates required fields

Frequently-asked questions

Have questions about media compression, what Ofinno is doing to shape it, or how your company can get involved? Here are answers to the ones we hear most.

What are MPEG and JVET?

MPEG (the Moving Picture Experts Group) is the standards body responsible for developing audio, video, and 3D media compression formats. JVET (the Joint Video Experts Team) is a collaboration between MPEG and the ITU’s video coding group that develops next-generation video codecs. Ofinno’s Advanced Media Lab contributes to both.

What is Ofinno’s role in media compression standards?

Ofinno is a top contributor to JVET and MPEG WG7. Our researchers are actively shaping the successor to VVC (H.266) — the next major video compression standard — while simultaneously contributing to standards for 3D point cloud compression, 3D mesh compression, and 3D Gaussian splatting.

What is the VVC successor and why does it matter?

VVC (Versatile Video Coding / H.266) is the current state-of-the-art video compression standard. Its successor, being developed within JVET, aims to achieve further significant improvements in compression efficiency — meaning higher quality video at lower bitrates. As video resolutions increase and streaming demand grows, each generation of codec directly impacts how efficiently content can be stored and delivered across networks worldwide.

What are point clouds, meshes, and Gaussian splatting?

These are three different ways of representing 3D content. Point clouds represent objects as collections of points in 3D space with associated attributes like color. Meshes represent surfaces using interconnected triangles, commonly used in animation and gaming. Gaussian splatting is a newer technique that represents scenes using overlapping 3D Gaussian functions, enabling fast, photorealistic rendering. All three need efficient compression standards to be practical for storage, transmission, and real-time delivery.

Can my company partner with Ofinno on media compression?

Yes. Ofinno works closely with partner companies, giving them first access to new inventions at a fraction of the cost of issued patents, with dedicated standards insight and prosecution support to grow portfolio value over time. Contact us to learn more.

Can my company license or acquire Ofinno’s media compression patents?

Yes. Ofinno licenses its standard essential patents under FRAND terms to companies whose products implement media compression standards, and selectively sells patent families to companies looking to strengthen their IP position. Reach out to discuss which option fits your needs.