Filer Named to Top 10 at NVIDIA Inception Grand Challenge APAC

Filer announced on the 22nd that it had been selected for the final top 10 at the APAC finals of the “NVIDIA Inception Grand Challenge 2026,” held in Singapore. More than 500 AI startups from the APAC region took part in this year’s competition, and Filer advanced to the final round after placing in the top 20, then made it into the final top 10 following its presentation review.

This year’s competition had no separate rankings, and the top 10 companies were named the final winners. Filer was recognized at an NVIDIA event for the second consecutive year, following its win at “AI Day Seoul” held in Korea last year.

The top 10 companies will receive opportunities to participate in NVIDIA global events, technical support, and follow-up proof-of-concept opportunities with partner companies.

“NVIDIA Inception Grand Challenge” is a program that expanded last year’s “AI Day Seoul,” held for Korean startups, across the APAC region. This year’s competition was run in a way that connected startups with real business challenges facing major companies and sought opportunities for technical validation, pilots, and business collaboration.

While working on advertising challenges for global FAST providers and CTV live broadcasting environments, Filer presented a technology in the Singapore final that increases ad inventory utilization based on scene-by-scene analysis of live video. The core of the technology lies in distinguishing video context in real time.

Filer said its video understanding engine analyzes dialogue, on-screen text, audio, brands, locations, and actions on a scene-by-scene basis. The engine is designed to determine the characteristics of each segment and whether it is safe.

Filer said it had extended its scene analysis technology, previously applied to VOD, to live streaming. Based on the results of this expansion, the company aims to lay the groundwork for ad placement based on scene-level context differentiation in news and live broadcasts as well.

Behind this are cases in some broadcast streaming environments where program-ad separation signals are lost, as well as the difficulty of distinguishing differences in value by scene. As a result, there are cases in which ad inventory is sold under the same standard. In addition, it is difficult to filter risky scenes in real time in news and live broadcasts, and such limits on risky-scene filtering can create constraints on ad operations.

Using “Video Understanding AI” that simultaneously analyzes video, audio, and text as its technological foundation, Filer analyzes content and context and operates by providing reference data for assessing content’s suitability for advertising and platform policies. The technology is applied in the areas of brand safety and T&S (Trust & Safety).

Filer performs scene-by-scene safety assessment and grading, enabling the conversion of live content into ad-eligible inventory. The company’s goal is to advance scene detection and link it to actual ad operations and brand safety management.

With this award, the company plans to push ahead with expanding its existing video understanding technology, verified on digital platforms, to CTV and live streaming. It also plans to broaden the scope of application for real-time video analysis technology in the global media and advertising markets.

Oh Jae-ho, CEO of Filer, said that last year’s domestic victory was a stage that proved the technology’s potential, and that this year’s APAC final was a moment that proved it could solve real industry issues. He added that the company will expand the application of its video understanding technology to CTV, live broadcasting, and global markets, and establish a trustworthy understanding layer.

Source: TECHWORLD · Kim Seung-ki
Original: https://www.epnc.co.kr/news/articleView.html?idxno=407400

References

• https://www.epnc.co.kr/news/articleView.html?idxno=407400