IBC: Bitmovin To Expand AI Scene Analysis (AISA) With AI Highlight Generation & Summarization Companion

Bitmovin, a provider of video streaming solutions, today rolled out “major updates” to its AI Scene Analysis (AISA) tool. Designed to help streaming providers unlock the full potential of their content libraries, the enhanced AI Scene Analysis now includes multi-language output, advanced scene-type detection for smarter ad placement, and upcoming AI-powered highlight generation for short-form clips.

With these updates, AI Scene Analysis “empowers broadcasters, OTT platforms, and content owners to automatically surface key video moments, generate AI-driven highlights, and create multi-language summaries — boosting audience engagement while driving new monetization opportunities,” according to Bitmovin.

Adding to the contextual advertising and auto ad scheduling pipelines announced in March, Bitmovin’s new short form clip generation feature automatically creates vertical videos, trailers and social media-ready clips from long-form content. This enables content providers to maximize content reuse and reach audiences across platforms like TikTok, Instagram Reels and YouTube Shorts.

The new multi-language output capability allows providers to create AI-powered summaries in multiple languages for their entire content catalog. “This not only enhances accessibility but also ensures content is discoverable across global audiences,” the company says.

Using AI Scene Analysis at the granular scene level, streaming services can:

• Improve monetization with context-aware ad placement.
• Avoid interrupting key moments by leveraging precise scene-boundary detection.
• Extract metadata aligned with IAB taxonomies to deliver more relevant, brand-safe advertising.

“Video providers are challenged with dealing with vast content libraries,” says Jacob Arends, Bitmovin senior product manager, playback and AI. “With these latest updates, AI Scene Analysis provides further value. Video providers can quickly create metadata and use it for recommendations and content reuse as well as batch processing for intelligent ad placement.”

Bitmovin also announced an upcoming Summarization Companion, which expands beyond simple summaries to deliver nuanced, verbose metadata. This metadata, it says, can power advanced large language model (LLM) workflows, improving content discovery, search, and personalized recommendations.

Additional innovations in the pipeline include voice-activated search to help developers locate code and documentation faster, and new agentic AI services for workflow automation.

Bitmovin will showcase the latest AI Scene Analysis pipelines in partner demos with Media Tailor, Broadpeak, Think Analysis, Data Graphs and Nomad at IBC 2025 (RAI Amsterdam, Sept. 12-15) at Booth 5.G58. Book a meeting here.