We bring you the best possible AI solutions for your company.

Fovea Super Resolution

Improve video perceptual quality with edge AI to premiumize user experiences.

Streaming Media Analytics

Measure user experience including bitrate, VMAF, frame drop rate, etc.

Low Latency Streaming Server

Enable peer-to-peer communication even in low-bandwidth regions.

Temporal Interpolation

Increase frame rate of video to improve perceptual quality (gaming and sports).

AR / VR Applications

Phoneme to viseme conversion on display interactions and communication.

Media Restoration

Restore classic SD content to HD/4K, remove artefacts, increase framerate to restore films.

Low Light Enhancement

Enhance the interpretability of an image captured with poor illumination.

Rotoscope Animation

Create an automatic matte outline for any element to enable intelligent video render

Language Transcription

Provide live transcription for low-resource languages.

Highlights Generation

Automated highlights for video content including for live broadcast.

Audio Content Recognition

Enable fast and accurate audio recognition and analysis on the local device itself.

Fovea STREAM Features.

Full-HD at SD bitrates

Content Agnostic

Client-side Integration

Encoding Profile Independent

Reduced CDN Cost

Improved Q.O.E.

QoE / QoS Analytics

Live, VOD & FAST Compatible

Industry Challenges.

…half of users (47%) claim to have unsubscribed from a streaming service due to buffering issues…

– Bitmovin, 2022, Buffering is costing streaming services, <Link>

…after just a 5 sec ad delay, 19.24% of viewers have already given up on waiting…

– Conviva, 2020, Benchmark the Performance of Every Stream, <Link>

…a single rebuffering event causes abandonment rate three times higher than a single bitrate change…

– Columbia University, IEEE INFOCOM 2016, QoE Matters More Than QoS: Why People Stop Watching Cat Videos, <Link>

…poor video quality and buffering could cost video streamers $390 million per month…

– Bitmovin, 2022, Buffering is costing streaming services, <Link>

Fovea STREAM Advantage

Real-time edge AI powered streaming with device-aware, content-aware, and context-ware enhancement.

  • Up to 30% reduction in rebuffering ratio*

  • Up to 30% bitrate switches* with no changes to streaming pipeline

  • Enable up to 40% savings+ on CDN costs

  • Enable HD/FHD experience for end users even in unreliable network conditions

* Actual improvement will depend on the ABR renditions for each application

+ ‘REAL-TIME PROCESSING OF MULTIMEDIA DATA ON AN EDGE DEVICE – Application no.202041023911, Status – Filed, Inventors – G. Suryanarayanan et.al

Demos.

Watch the demo video to see the Fovea STREAM suite of solutions for real-time Edge AI Video Enhancement in action.

Input 180p – Fovea STREAM Output 1080p

Original HD – Fovea STREAM Equivalent 480p

Case Studies.

Indian top 5 UGC platform sees 40% CDN cost saving

Exponential increase in the number of subscribers to the UGC app, both to produce and consume content, rapidly increased challenges. Owing to the variability in cell phones and conditions used by content creators or initial video quality for content, which was reproduced, the video quality varied significantly within each video for this platform.

CDN Cost Saved
Average Bitrate Saved
VMAF Points Improved

Whitepapers.

Video Encoding: Codecs

Learn about the different types of codecs and the advantages and disadvantages of using each.
Read More

Video Encoding: Rate Control Mode

Learn about the different types of bitrate, their impact on quality and recommendations for best outcomes.

Read More

Compare Fovea STREAM.

Significant reduction in data costs borne by OTT and premiumization of content for UGC.

  • Works real-time on the customer edge device / smartphone chip-set.
  • Works on any input resolution and any quality.
  • Works irrespective of the encoding format.
  • Works with absolutely no latency, on any type of media content.
  • Works for all kind of video streaming, VOD and Live formats.
  • Works on smartphones, smart-TVs, tablets, PC; its device agnostic technology.

Current media streaming AI solutions

  • Often tends to be offline GPU / cloud based processing, streamed on the edge device.
  • Depends on the resolution provided by the content provider.
  • Latency can depend on the network bandwidth, kind of media being streamed.
  • Encoding parameters decide compatible target devices and server availability.
  • Currently can only work for VOD and not for live streaming content.

  • Depends heavily on the kind of device used for streaming.

Need a personalized solution?

Customize AI development to target the exact specifications and requirements that your business needs

Get a personal consultation.

Office

A-202/203, Miraya Rose, Siddapura, Whitefield, Bengaluru, 560066

Contact

social@www.myelinfoundry.com

Open Hours

Monday to Friday: 8am – 6pm