Training Noisy Single-Channel Speech Separation With Noisy Oracle Sources: A Large Gap and A Small Step
This addresses the challenge of training speech separation models with real-world noisy data for applications in audio processing, but it is incremental as it builds on existing synthetic mixture approaches.
The paper tackles the problem of training single-channel speech separation systems in noisy conditions, where using noisy oracle sources leads to performance degradation due to noise inseparability; it proposes a new training objective that exploits this inseparability to improve separation, achieving a 1.2 dB SI-SDR gain over baseline methods.
As the performance of single-channel speech separation systems has improved, there has been a desire to move to more challenging conditions than the clean, near-field speech that initial systems were developed on. When training deep learning separation models, a need for ground truth leads to training on synthetic mixtures. As such, training in noisy conditions requires either using noise synthetically added to clean speech, preventing the use of in-domain data for a noisy-condition task, or training using mixtures of noisy speech, requiring the network to additionally separate the noise. We demonstrate the relative inseparability of noise and that this noisy speech paradigm leads to significant degradation of system performance. We also propose an SI-SDR-inspired training objective that tries to exploit the inseparability of noise to implicitly partition the signal and discount noise separation errors, enabling the training of better separation systems with noisy oracle sources.