This paper introduces a speech enhancement solution tailored for true wireless stereo (TWS) earbuds on-device usage.The solution was specifically designed to support conversations in noisy environments, with active noise cancellation (ANC) activated.The primary challenges for speech enhancement models in this context arise from computational complexity that limits on-device usage and latency that must be less than 3 ms to preserve a live conversation.To address these issues, we evaluated several crucial design elements, including the network architecture and domain, design of loss functions, pruning method, and hardware-specific optimization.Consequently, we demonstrated substantial improvements in speech enhancement quality compared with that in baseline models, while simultaneously reducing the computational complexity and algorithmic latency.

In this paper, we address the challenge of speech enhancement in real-world recordings, which often contain various forms of distortion, such as background noise, reverberation, and microphone artifacts. We revisit the use of Generative Adversarial Networks (GANs) for speech enhancement and theoretically show that GANs are naturally inclined to seek the point of maximum density within the conditional clean speech distribution, which, as we argue, is essential for speech enhancement task. We study various feature extractors for perceptual loss to facilitate the stability of adversarial training, developing a methodology for probing the structure of the feature space. This leads us to integrate WavLM-based perceptual loss into MS-STFT adversarial training pipeline, creating an effective and stable training procedure for the speech enhancement model. The resulting speech enhancement model, which we refer to as FINALLY, builds upon the HiFi++ architecture, augmented with a WavLM encoder and a novel training pipeline. Empirical results on various datasets confirm our model's ability to produce clear, high-quality speech at 48 kHz, achieving state-of-the-art performance in the field of speech enhancement. Demo page: https://samsunglabs.github.io/FINALLY-page/

This study focuses on understanding the purchase decision-making process of B2B buyers in the context of inside sales. While many studies have explored this topic in a B2C context, there has been little attention given to the unique features of B2B inside sales. To address this gap, we developed and empirically validated a buyer's intention to purchase (BIP) model that integrates the B2B purchase decision-making process and buying behavior theories. Using PLS-SEM analysis on data collected from 126 B2B buyers, we found that trust in sellers and connection flexibility are crucial factors in determining buyers' intention to purchase. Specifically, these factors influence the purchase intention via appointment and contact willingness during different phases of the B2B purchase process. Additionally, our multigroup analyses suggest that buyers' demographic and firmographic attributes affect the purchase decision-making process in different ways. By identifying different groups of buyers and their distinctive needs and preferences, our findings inform more effective approaching and selling strategies. This study contributes to both academic and sales practices by providing critical insights into the factors that influence the B2B inside sales process, which ultimately improve sales performance.

Generative adversarial networks have recently demonstrated outstanding performance in neural vocoding outperforming best autoregressive and flow-based models. In this paper, we show that this success can be extended to other tasks of conditional audio generation. In particular, building upon HiFi vocoders, we propose a novel HiFi++ general frame-work for bandwidth extension and speech enhancement. We show that with the improved generator architecture, HiFi++ performs better or comparably with the state-of-the-art in these tasks while spending significantly less computational resources. The effectiveness of our approach is validated through a series of extensive experiments.

Generative adversarial networks (GANs) have an enormous potential impact on digital content creation, e.g., photorealistic digital avatars, semantic content editing, and quality enhancement of speech and images. However, the performance of modern GANs comes together with massive amounts of computations performed during the inference and high energy consumption. That complicates, or even makes impossible, their deployment on edge devices. The problem can be reduced with quantization—a neural network compression technique that facilitates hardware-friendly inference by replacing floating-point computations with low-bit integer ones. While quantization is well established for discriminative models, the performance of modern quantization techniques in application to GANs remains unclear. GANs generate content of a more complex structure than discriminative models, and thus quantization of GANs is significantly more challenging. To tackle this problem, we perform an extensive experimental study of state-of-art quantization techniques on three diverse GAN architectures, namely StyleGAN, Self-Attention GAN, and CycleGAN. As a result, we discovered practical recipes that allowed us to successfully quantize these models for inference with 4/8-bit weights and 8-bit activations while preserving the quality of the original full-precision models.

It is predicted that significant and ongoing investment in M-Commerce platforms and application development by commercial entities will fundamentally change consumers' shopping and web browsing behaviours. However, the evolving behaviour of Smart Phone users is somewhat tempered by concerns over M-Payments. If Smart Phones are to reach their full M-Commerce potential, the ability of consumers to transact and pay for products/services through these devices in an easy, safe and reliable manner must be addressed. In response, this paper contributes a theoretical model and empirically tests the model to explore Irish consumers' perceptions of using Smart Phones to make M-Payments for products/services. The findings present conclusive evidence that trust is the most powerful factor influencing consumers' willingness to use Smart Phones to make M-Payments. While perceived usefulness and perceived ease of use influence the payment decision, their impact is much lower. Mobile self-efficacy and personal innovativeness have almost no direct impact. The paper concludes that irrespective of individuals' high levels of personal innovativeness or mobile self-efficacy and irrespective of whether Smart Mobile Media Services are perceived as useful and easy to use, consumers will not make M-Payments, until they are convinced that Smart Phone M-Payment systems are safe and reliable.