International Audio Laboratories Erlangen

Demos

Acoustic Scene Analysis

• Python-Based Real-Time Direction-of-Arrival Estimation Using Deep Learning

• Multi-Source and Signal-Aware DOA Estimation Using DNNs

• Deep Learning-based Reverberation Time and Direct-to-Reverberation Ratio Estimation

• End-to-End Signal-Aware Direction-of-Arrival Estimation Using Weighted Steered-Response Power

Artificial Reverberation and Reverberation Enhancement

• Optimized Velvet-Noise Decorrelator

• Sign-Agnostic Matrix Design for Spatial Artificial Reverberation with Feedback Delay Networks

• Modal Decomposition of Feedback Delay Networks

• Dense Reverberation with Delay Feedback Matrices

• Reverberation Enhancement Systems

• Neural Audio Decorrelation Using Generative Adversarial Networks (NEW)

• Neural Binaural Upmix of Stereo Content (NEW)

• Matching Reverberant Speech Through Learned Acoustic Embeddings and Feedback Delay Networks (NEW)

Distributed Array Processing

• Geometry-based Spatial Sound Acquisition Using Distributed Microphone Arrays

• Informed Spatial Filtering for Sound Extraction Using Distributed Microphone Arrays

• Speech Enhancement with a Low-complexity Online Source Number Estimator Using Distributed Arrays

• Near-field Source Extraction Using Speech Presence Probabilities for Ad-Hoc Microphone Arrays

• Spotforming: Spatial Filtering With Distributed Arrays for Position-Selective Sound Acquisition

• Blind Source Separation of Moving Sources Using Sparsity-Based Source Detection and Tracking

• Sample Rate Offset Compensated Acoustic Echo Cancellation for Multi-Device Scenarios (NEW)

Spatial Audio Processing

• Subband Center Scaling using Power Ratios

Spatial Sound Acquisition

• Extracting Reverberant Sound Using a Linearly Constrained Minimum Variance Spatial Filter

• An Informed Parametric Spatial Filter Based on Instantaneous Direction-of-Arrival Estimates

• Automatic Spatial Gain Control for an Informed Spatial Filter

• An Acoustical Zoom Based on Informed Spatial Filtering

• Direction Preserving Wiener Matrix Filtering for Ambisonic Input-Output Systems

• Direction and Reverberation Preserving Noise Reduction of Ambisonics Signals

• Neural Directional Filtering - Far-Field Directivity Control with a Small Microphone Array (NEW)

• Neural Direction Filtering (NEW)

• Neural Directional Filtering with Configurable Directivity Pattern at Inference (NEW)

Speaker Separation and Extraction

• Signal-Guided Source Separation

• Video-based Guided Source Separation

• Speech Separation for an Unknown Number of Speakers Using Transformers With Encoder-Decoder Attractors (NEW)

• Multi-Microphone Speaker Separation by Spatial Regions (NEW)

• Speaker Extraction for Multi-Party Communication Applications (NEW)

• New Insights on Target Speaker Extraction (NEW)

Speech Coding

• Acoustic Teleportation via Disentangled Neural Audio Codec Representations (NEW)

Speech Enhancement (Directional Filtering, Noise Reduction, Dereverberation)

• An Informed Spatial Filter for Dereverberation in the Spherical Harmonic Domain

• Multichannel Dereverberation for Hearing Aids With Interaural Coherence Preservation

• Relative Transfer Function Estimation Exploiting Instantaneous Signals and the Signal Subspace

• Minimum Bayes-Risk Signal Detection for Speech Enhancement Based on a Narrowband DOA Model

• Single-Channel Dereverberation Using Relative Convolutive Transfer Functions

• Head-Orientation Compensation With Video-Informed Single Channel Speech Enhancement

• Online Dereverberation for Dynamic Scenarios Using a Kalman Filter With an Autoregressive Model

• A Bayesian Approach to Informed Spatial Filtering With Robustness Against DOA Estimation Errors

• Non-Stationary Noise PSD Matrix Estimation for Multi-Channel Blind Speech Extraction

• Linear Prediction Based Online Dereverberation and Noise Reduction Using Alternating Kalman Filters

• Evaluation and Comparison of Late Reverberation Power Spectral Density Estimators

• Classification vs. Regression in Supervised Learning for Single Channel Speaker Count Estimation

• Joint Noise Reduction and Dereverberation

• Deep Filtering: Signal Extraction and Reconstruction Using Complex Time-Frequency Filters

• Time-Frequency Masking Based Online Speech Enhancement With Multi-Channel Data Using Convolutional Neural Networks

• Time-Frequency Masking Based Online Multi-Channel Speech Enhancement With Convolutional Recurrent Neural Networks

• Multi-Channel Wind Noise Reduction Using the Corcos Model

• Spatial Coherence-Aware Multi-Channel Wind Noise Reduction

• Declipping Speech Using Deep Filtering

• Signal-Dependent Mixing for Direction-Preserving Multichannel Noise Reduction

• Improved Normalizing Flow-Based Speech Enhancement using an All-pole Gammatone Filterbank for Conditional Input Representation (NEW)

• Dynamic Slimmable Networks for Efficient Speech Separation (NEW)

Speech Synthesis and Speaker Anonymization

• Multi-Speaker Text-to-Speech Using ForwardTacotron with Improved Duration Prediction (NEW)

• Subjective Evaluation of Text-to-Speech Models: Comparing Absolute Category Rating and Ranking by Elimination Tests (NEW)

• Target Speaker Adaptation in Text-to-Speech Synthesis: A Comparison of Efficient Fine-Tuning and Zero-Shot Methods (NEW)

• Low-Resource Text-to-Speech Synthesis Using Noise-Augmented Training of ForwardTacotron (NEW)