숏폼 콘텐츠 홍수 속에서 '약시'를 VR로 재정의한 Amblyotube의 기술적 접근법 ㅤ 인도의 마이크로드라마 플랫폼 EloElo가 NBCUniversal로부터 8,000만 달러를 유치하며, 숏폼 비디오 시장의 폭발적 성장을…

Amblyotube, a VR platform aiming to redefine short-form content, tackles a surprising paradox: modern high-resolution displays and refresh rates are driving prolonged close-range screen time, which can exacerbate amblyopia and visual impairment, particularly in developing adolescents. To address this, Amblyotube tackles the issue at the hardware level by utilizing Meta Quest 2/3/Pro/3S's dual-display architecture to split a single YouTube stream into a dichoptic pair during runtime.

The core pipeline consists of three stages: 1) Dominant Eye Shader (digital amblyopia) which applies a Gaussian blur, contrast/gamma/invert shader to the dominant eye frame buffer, resulting in a fully opaque output but at 30-60% transparency during real-world use to encourage simultaneous binocular processing. 2) Lazy Eye Sharpener (MFBF mode) uses MediaPipe for real-time pose estimation to extract the ROI of the amblyopic eye text, applying unsharp masking and high-pass filter chains to enhance visual clarity.

3) Magenta Focus Cue + Wedge Occlusion, which renders the magenta (#FF00FF) target for the amblyopic eye and neutral gray for the dominant eye, physically blocking binocular rivalry regions with a wedge mask. Initial settings are automatically adjusted using the LogMAR staircase method (30 trials, 3m virtual distance, Landolt C) and mapped to shader uniform presets for pre-set levels of difficulty.

The platform supports pass-through AR mode, maintaining real-world overlays by adjusting composition layer orders. Sessions range from 30-40 minutes per day, up to a 1-hour limit, avoiding medical device certification but positioning as a supplementary training tool with Unity/OpenXR native implementation guaranteeing 72fps stable frames.