Optimizing Video with AI: The VisualOn Optimizer Explained

Here I spoke with Yang Cai, President & CEO at VisualOn, Inc. about the AI-powered VisualOn Optimizer, which can reduce bandwidth consumption by as much as 70% without compromising video quality. As Yang explained, the Optimizer analyzes video content before transcoding and adjusts the bitrate dynamically to maintain the target quality. It also provides an optional preprocessing feature that can improve source quality before transcoding.

Contents:

00:00:11 - Optimizer Functionality - Yang Cai explains how the optimizer works with encoders and its flexibility across different codecs.

00:00:49 - Codec Integration and Flexibility - Discussion on the optimizer's compatibility with a wide range of encoders and compression formats.

00:01:39 - Bandwidth Savings and Efficiency - Insights into the potential bandwidth savings and how these are influenced by content complexity.

00:02:10 - Pricing Model - The optimizer's cost-saving benefits and pricing structure based on customer savings.

00:02:35 - AI Integration and Enhancement - The addition of AI technology to the optimizer and its impact on performance.

00:03:29 - Dynamic Parameter Adjustment - How the optimizer dynamically adjusts encoding parameters on a frame-by-frame basis.

00:03:53 - Quality Optimization Strategies - The approach to maintaining target quality levels through selective parameter adjustments.

00:04:47 - AI Training and Quality Assurance - The methodology behind training the optimizer's AI and ensuring quality.

00:05:21 - Metrics and Measurement - How different metrics like PSNR and SSIM compare when used with the optimizer.

00:05:48 - Customer Validation and Results - External validation of the optimizer's performance in real-world applications.

00:06:18 - Encoder Integration - The process of integrating the optimizer with various encoding platforms.

00:07:15 - Preprocessing and Pixel Adjustment - The optional preprocessing feature and its impact on encoding efficiency and video quality.

00:07:52 - HDR Content Handling - The optimizer's approach to HDR content and maintaining quality in HDR video streams.

10:21:08 - Trust and Verification - Addressing concerns about the reliability of VMAF and exploring the optimizer's performance with other metrics like PSNR and SSIM.

11:17:19 - Real-World Validation - Insights into the optimizer's effectiveness through customer case studies and real-world production environments.

12:16:07 - Integration Challenges and Solutions - Discussing the integration of the optimizer with various encoding platforms, including proprietary and non-FFmpeg based systems.

13:50:09 - Scalability and Efficiency in Large-Scale Transcoding - Addressing the optimizer's performance and scalability in extensive transcoding setups.

16:13:05 - Preprocessing and Content Adjustment - The quality impact of the optional preprocessing feature.

17:17:00 - Customization and User Experience - Discussing the flexibility offered to users in terms of customization and the ease of integration with the optimizer SDK.

18:26:18 - Creative Intent and Quality Control - Considering the balance between compression efficiency and maintaining the creative intent of the original content.

About This Video:

This was one of the interviews recorded to produce this session at the NAB 2024 Streaming Summit - AI TRACK: Beyond the Hype: A Critical Look at AI in Video Streaming
STREAMING SUMMIT.

Monday, April 15 | 1:30 PM - 2:00 PM PST/Pass Required: Streaming Summit Pass

Description: Artificial intelligence has become a buzzword in all technologies, including streaming video, with the promise of revolutionizing how we create, compress, and watch videos. However, it's crucial to separate the hype from reality. In this session, Jan Ozer from the Streaming Learning Center explores the most recent streaming-related AI announcements to detail which to dismiss, which will become relevant, and when.

https://nab24.mapyourshow.com/8_0/ses...