Researchers have developed a novel hybrid time-spectrogram transformer architecture to address the challenge of separating distinct noise sources within electric vehicle interiors. This method aims to decompose recorded interior sounds into components like wind noise, tire noise, and powertrain noise, which is crucial for targeted acoustic engineering. The model is trained on synthesized data from a physics-based generator and incorporates operating-state information to improve separation accuracy, offering a feasible approach for analyzing acoustic contributions in technical systems. AI
IMPACT Could enable more precise acoustic engineering in vehicles and other technical systems by accurately identifying noise sources.
RANK_REASON Academic paper detailing a new AI model architecture for a specific technical problem. [lever_c_demoted from research: ic=1 ai=1.0]
- arXiv
- auxiliary-drive noise
- Electric drive unit cooling and lubrication system with bearing shims and rotor shaft channel
- electric vehicle
- Hugging Face
- inverter
- powertrain noise
- Tire noise reduction device and pneumatic tire provided with same
- Transfer path analysis of a railway vehicle based on global transfer direct transfer
- transformer
- wind noise
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
