FLowHigh: Towards Efficient and High-Quality Audio Super-Resolution with Single-Step Flow Matching

Audio super-resolution is challenging owing to its ill-posed nature. Recently, the application of diffusion models in audio super-resolution has shown promising results in alleviating this challenge. However, diffusion-based models have limitations, primarily the necessity for numerous sampling steps, which causes significantly increased latency when synthesizing high-quality audio samples. In this paper, we propose FLowHigh, a novel approach that integrates flow matching, a highly efficient generative model, into audio super-resolution. We also explore probability paths specially tailored for audio super-resolution, which effectively capture high-resolution audio distributions, thereby enhancing reconstruction quality. The proposed method generates high-fidelity, high-resolution audio through a single-step sampling process across various input sampling rates. The experimental results on the VCTK benchmark dataset demonstrate that FLowHigh achieves state-of-the-art performance in audio super-resolution, as evaluated by log-spectral distance and ViSQOL while maintaining computational efficiency with only a single-step sampling process.

AEROMamba: An efficient architecture for audio super-resolution using generative adversarial networks and state space models

Audio super-resolution aims to enhance low-resolution signals by creating high-frequency content. In this work, we modify the architecture of AERO (a state-of-the-art system for this task) for music super-resolution. SPecifically, we replace its original Attention and LSTM layers with Mamba, a State Space Model (SSM), across all network layers. Mamba is capable of effectively substituting the mentioned modules, as it offers a mechanism similar to that of Attention while also functioning as a recurrent network. With the proposed AEROMamba, training requires 2-4x less GPU memory, since Mamba exploits the convolutional formulation and leverages GPU memory hierarchy. Additionally, during inference, Mamba operates in constant memory due to recurrence, avoiding memory growth associated with Attention. This results in a 14x speed improvement using 5x less GPU. Subjective listening tests (0 to 100 scale) show that the proposed model surpasses the AERO model. In the MUSDB dataset, degraded signals scored 38.22, while AERO and AEROMamba scored 60.03 and 66.74, respectively. For the PianoEval dataset, scores were 72.92 for degraded signals, 76.89 for AERO, and 84.41 for AEROMamba.

TRAMBA: A Hybrid Transformer and Mamba Architecture for Practical Audio and Bone Conduction Speech Super Resolution and Enhancement on Mobile and Wearable Platforms

We propose TRAMBA, a hybrid transformer and Mamba architecture for acoustic and bone conduction speech enhancement, suitable for mobile and wearable platforms. Bone conduction speech enhancement has been impractical to adopt in mobile and wearable platforms for several reasons: (i) data collection is labor-intensive, resulting in scarcity; (ii) there exists a performance gap between state of-art models with memory footprints of hundreds of MBs and methods better suited for resource-constrained systems. To adapt TRAMBA to vibration-based sensing modalities, we pre-train TRAMBA with audio speech datasets that are widely available. Then, users fine-tune with a small amount of bone conduction data. TRAMBA outperforms state-of-art GANs by up to 7.3% in PESQ and 1.8% in STOI, with an order of magnitude smaller memory footprint and an inference speed up of up to 465 times. We integrate TRAMBA into real systems and show that TRAMBA (i) improves battery life of wearables by up to 160% by requiring less data sampling and transmission; (ii) generates higher quality voice in noisy environments than over-the-air speech; (iii) requires a memory footprint of less than 20.0 MB.

Gull: A Generative Multifunctional Audio Codec

We introduce Gull, a generative multifunctional audio codec. Gull is a general purpose neural audio compression and decompression model which can be applied to a wide range of tasks and applications such as real-time communication, audio super-resolution, and codec language models. The key components of Gull include (1) universal-sample-rate modeling via subband modeling schemes motivated by recent progress in audio source separation, (2) gain-shape representations motivated by traditional audio codecs, (3) improved residual vector quantization modules for simpler training, (4) elastic decoder network that enables user-defined model size and complexity during inference time, (5) built-in ability for audio super-resolution without the increase of bitrate. We compare Gull with existing traditional and neural audio codecs and show that Gull is able to achieve on par or better performance across various sample rates, bitrates and model complexities in both subjective and objective evaluation metrics.

Low-power SNN-based audio source localisation using a Hilbert Transform spike encoding scheme

Sound source localisation is used in many consumer electronics devices, to help isolate audio from individual speakers and to reject noise. Localization is frequently accomplished by "beamforming" algorithms, which combine microphone audio streams to improve received signal power from particular incident source directions. Beamforming algorithms generally use knowledge of the frequency components of the audio source, along with the known microphone array geometry, to analytically phase-shift microphone streams before combining them. A dense set of band-pass filters is often used to obtain known-frequency "narrowband" components from wide-band audio streams. These approaches achieve high accuracy, but state of the art narrowband beamforming algorithms are computationally demanding, and are therefore difficult to integrate into low-power IoT devices. We demonstrate a novel method for sound source localisation in arbitrary microphone arrays, designed for efficient implementation in ultra-low-power spiking neural networks (SNNs). We use a novel short-time Hilbert transform (STHT) to remove the need for demanding band-pass filtering of audio, and introduce a new accompanying method for audio encoding with spiking events. Our beamforming and localisation approach achieves state-of-the-art accuracy for SNN methods, and comparable with traditional non-SNN super-resolution approaches. We deploy our method to low-power SNN audio inference hardware, and achieve much lower power consumption compared with super-resolution methods. We demonstrate that signal processing approaches can be co-designed with spiking neural network implementations to achieve high levels of power efficiency. Our new Hilbert-transform-based method for beamforming promises to also improve the efficiency of traditional DSP-based signal processing.

Real3D-Portrait: One-shot Realistic 3D Talking Portrait Synthesis

One-shot 3D talking portrait generation aims to reconstruct a 3D avatar from an unseen image, and then animate it with a reference video or audio to generate a talking portrait video. The existing methods fail to simultaneously achieve the goals of accurate 3D avatar reconstruction and stable talking face animation. Besides, while the existing works mainly focus on synthesizing the head part, it is also vital to generate natural torso and background segments to obtain a realistic talking portrait video. To address these limitations, we present Real3D-Potrait, a framework that (1) improves the one-shot 3D reconstruction power with a large image-to-plane model that distills 3D prior knowledge from a 3D face generative model; (2) facilitates accurate motion-conditioned animation with an efficient motion adapter; (3) synthesizes realistic video with natural torso movement and switchable background using a head-torso-background super-resolution model; and (4) supports one-shot audio-driven talking face generation with a generalizable audio-to-motion model. Extensive experiments show that Real3D-Portrait generalizes well to unseen identities and generates more realistic talking portrait videos compared to previous methods. Video samples and source code are available at https://real3dportrait.github.io .

Generative AI in Vision: A Survey on Models, Metrics and Applications

Generative AI models have revolutionized various fields by enabling the creation of realistic and diverse data samples. Among these models, diffusion models have emerged as a powerful approach for generating high-quality images, text, and audio. This survey paper provides a comprehensive overview of generative AI diffusion and legacy models, focusing on their underlying techniques, applications across different domains, and their challenges. We delve into the theoretical foundations of diffusion models, including concepts such as denoising diffusion probabilistic models (DDPM) and score-based generative modeling. Furthermore, we explore the diverse applications of these models in text-to-image, image inpainting, and image super-resolution, along with others, showcasing their potential in creative tasks and data augmentation. By synthesizing existing research and highlighting critical advancements in this field, this survey aims to provide researchers and practitioners with a comprehensive understanding of generative AI diffusion and legacy models and inspire future innovations in this exciting area of artificial intelligence.

AudioSR: Versatile Audio Super-resolution at Scale

Audio super-resolution is a fundamental task that predicts high-frequency components for low-resolution audio, enhancing audio quality in digital applications. Previous methods have limitations such as the limited scope of audio types (e.g., music, speech) and specific bandwidth settings they can handle (e.g., 4kHz to 8kHz). In this paper, we introduce a diffusion-based generative model, AudioSR, that is capable of performing robust audio super-resolution on versatile audio types, including sound effects, music, and speech. Specifically, AudioSR can upsample any input audio signal within the bandwidth range of 2kHz to 16kHz to a high-resolution audio signal at 24kHz bandwidth with a sampling rate of 48kHz. Extensive objective evaluation on various audio super-resolution benchmarks demonstrates the strong result achieved by the proposed model. In addition, our subjective evaluation shows that AudioSR can acts as a plug-and-play module to enhance the generation quality of a wide range of audio generative models, including AudioLDM, Fastspeech2, and MusicGen. Our code and demo are available at https://audioldm.github.io/audiosr.

BAE-Net: A Low complexity and high fidelity Bandwidth-Adaptive neural network for speech super-resolution

Speech bandwidth extension (BWE) has demonstrated promising performance in enhancing the perceptual speech quality in real communication systems. Most existing BWE researches primarily focus on fixed upsampling ratios, disregarding the fact that the effective bandwidth of captured audio may fluctuate frequently due to various capturing devices and transmission conditions. In this paper, we propose a novel streaming adaptive bandwidth extension solution dubbed BAE-Net, which is suitable to handle the low-resolution speech with unknown and varying effective bandwidth. To address the challenges of recovering both the high-frequency magnitude and phase speech content blindly, we devise a dual-stream architecture that incorporates the magnitude inpainting and phase refinement. For potential applications on edge devices, this paper also introduces BAE-NET-lite, which is a lightweight, streaming and efficient framework. Quantitative results demonstrate the superiority of BAE-Net in terms of both performance and computational efficiency when compared with existing state-of-the-art BWE methods.

Phase Repair for Time-Domain Convolutional Neural Networks in Music Super-Resolution

Audio Super-Resolution (SR) is an important topic in the field of audio processing. Many models are designed in time domain due to the advantage of waveform processing, such as being able to avoid the phase problem. However, in prior works it is shown that Time-Domain Convolutional Neural Network (TD-CNN) approaches tend to produce annoying artifacts in their output. In order to confirm the source of the artifact, we conduct an AB listening test and found phase to be the cause. We further propose Time-Domain Phase Repair (TD-PR) to improve TD-CNNs' performance by repairing the phase of the TD-CNNs' output. In this paper, we focus on the music SR task, which is challenging due to the wide frequency response and dynamic range of music. Our proposed method can handle various narrow-bandwidth from 2.5kHz to 4kHz with a target bandwidth of 8kHz. We conduct both objective and subjective evaluation to assess the proposed method. The objective evaluation result indicates the proposed method achieves the SR task effectively. Moreover, the proposed TD-PR obtains the much higher mean opinion scores than all TD-CNN baselines, which indicates that the proposed TD-PR significantly improves perceptual quality. Samples are available on the demo page.