Towards Context-aware Automatic Haptic Effect Generation for Home Theatre Environments
summary
An automated pipeline extracts four psychoacoustic features (sharpness, booming, low‑frequency energy, loudness) and CNN visual saliency (VGG backbone with spatial and channel attention) to map events to a chair instrumented with nine actuators and phantom‑sensation rendering. The algorithm thresholds and fuses signals to set spatiotemporal amplitudes; a user study ( n=16 ) across six clips finds the multimodal condition outperforms audio‑only, visual‑only and random (ANOVA significant). Limitations include dark or soundless scenes where saliency or audio cues fail.
AI-written summary · about
chapters (11)
- 0:00Intro
- 0:294D Movie Chair
- 1:45Algorithm Overview
- 3:35Visual Modality Analysis
- 4:51Information Fusion and Hardware Implementation
- 6:16User Study
- 7:13Questionnaire
- 8:06Two-way ANOVA Results
- 8:50Expandability of the Algorithm
- 9:16Extensibility of the Algorithm
- 9:36Takeaway
related talks
09mMulti-modal Multi-scale Attention Guidance in Cyber-Physical Environments
13mEnhancing XR Audio Realism via Multimodal Scene-Aware Acoustic Rendering
42mFeeling Films – The Future of Tactile Media
07mWeirding Haptics: In-Situ Prototyping of Vibrotactile Feedback in Virtual Reality through ...
13mAttraCar: Multisensory In-Car VR with Thermal, Airflow, and Motion Feedback through Built-In Vehi...
06mCoolMoves: User Motion Accentuation in Virtual Reality
09mThe Subconscious Director: Dynamically Personalizing Videos Using Gaze Data
20mExploring the hidden potential of sound data