Ahmed Mustafa

Ahmed Mustafa, Jean-Marc Valin, Jan Büthe et al.

GAN vocoders are currently one of the state-of-the-art methods for building high-quality neural waveform generative models. However, most of their architectures require dozens of billion floating-point operations per second (GFLOPS) to generate speech waveforms in samplewise manner. This makes GAN vocoders still challenging to run on normal CPUs without accelerators or parallel computers. In this work, we propose a new architecture for GAN vocoders that mainly depends on recurrent and fully-connected networks to directly generate the time domain signal in framewise manner. This results in considerable reduction of the computational cost and enables very fast generation on both GPUs and low-complexity CPUs. Experimental results show that our Framewise WaveGAN vocoder achieves significantly higher quality than auto-regressive maximum-likelihood vocoders such as LPCNet at a very low complexity of 1.2 GFLOPS. This makes GAN vocoders more practical on edge and low-power devices.

Ahmed Mustafa, Muhammad Tahir Rafique, Muhammad Ijlal Baig et al.

This research paper introduces a novel word-level Optical Character Recognition (OCR) model specifically designed for digital Urdu text, leveraging transformer-based architectures and attention mechanisms to address the distinct challenges of Urdu script recognition, including its diverse text styles, fonts, and variations. The model employs a permuted autoregressive sequence (PARSeq) architecture, which enhances its performance by enabling context-aware inference and iterative refinement through the training of multiple token permutations. This method allows the model to adeptly manage character reordering and overlapping characters, commonly encountered in Urdu script. Trained on a dataset comprising approximately 160,000 Urdu text images, the model demonstrates a high level of accuracy in capturing the intricacies of Urdu script, achieving a CER of 0.178. Despite ongoing challenges in handling certain text variations, the model exhibits superior accuracy and effectiveness in practical applications. Future work will focus on refining the model through advanced data augmentation techniques and the integration of context-aware language models to further enhance its performance and robustness in Urdu text recognition.

Muhammad Tahir Rafique, Ahmed Mustafa, Hasan Sajid

The growing demand for road use in urban areas has led to significant traffic congestion, posing challenges that are costly to mitigate through infrastructure expansion alone. As an alternative, optimizing existing traffic management systems, particularly through adaptive traffic signal control, offers a promising solution. This paper explores the use of Reinforcement Learning (RL) to enhance traffic signal operations at intersections, aiming to reduce congestion without extensive sensor networks. We introduce two RL-based algorithms: a turn-based agent, which dynamically prioritizes traffic signals based on real-time queue lengths, and a time-based agent, which adjusts signal phase durations according to traffic conditions while following a fixed phase cycle. By representing the state as a scalar queue length, our approach simplifies the learning process and lowers deployment costs. The algorithms were tested in four distinct traffic scenarios using seven evaluation metrics to comprehensively assess performance. Simulation results demonstrate that both RL algorithms significantly outperform conventional traffic signal control systems, highlighting their potential to improve urban traffic flow efficiently.

Wolfgang Mack, Ahmed Mustafa, Rafał Łaganowski et al.

Neural codecs, comprising an encoder, quantizer, and decoder, enable signal transmission at exceptionally low bitrates. Training these systems requires techniques like the straight-through estimator, soft-to-hard annealing, or statistical quantizer emulation to allow a non-zero gradient across the quantizer. Evaluating the effect of quantization in neural codecs, like the influence of gradient passing techniques on the whole system, is often costly and time-consuming due to training demands and the lack of affordable and reliable metrics. This paper proposes an efficient evaluation framework for neural codecs using simulated data with a defined number of bits and low-complexity neural encoders/decoders to emulate the non-linear behavior in larger networks. Our system is highly efficient in terms of training time and computational and hardware requirements, allowing us to uncover distinct behaviors in neural codecs. We propose a modification to stabilize training with the straight-through estimator based on our findings. We validate our findings against an internal neural audio codec and against the state-of-the-art descript-audio-codec.

Ahmed Mustafa, Jan Büthe, Srikanth Korse et al.

Recently, GAN vocoders have seen rapid progress in speech synthesis, starting to outperform autoregressive models in perceptual quality with much higher generation speed. However, autoregressive vocoders are still the common choice for neural generation of speech signals coded at very low bit rates. In this paper, we present a GAN vocoder which is able to generate wideband speech waveforms from parameters coded at 1.6 kbit/s. The proposed model is a modified version of the StyleMelGAN vocoder that can run in frame-by-frame manner, making it suitable for streaming applications. The experimental results show that the proposed model significantly outperforms prior autoregressive vocoders like LPCNet for very low bit rate speech coding, with computational complexity of about 5 GMACs, providing a new state of the art in this domain. Moreover, this streamwise adversarial vocoder delivers quality competitive to advanced speech codecs such as EVS at 5.9 kbit/s on clean speech, which motivates further usage of feed-forward fully-convolutional models for low bit rate speech coding.

Ahmed Mustafa, Nicola Pia, Guillaume Fuchs

In recent years, neural vocoders have surpassed classical speech generation approaches in naturalness and perceptual quality of the synthesized speech. Computationally heavy models like WaveNet and WaveGlow achieve best results, while lightweight GAN models, e.g. MelGAN and Parallel WaveGAN, remain inferior in terms of perceptual quality. We therefore propose StyleMelGAN, a lightweight neural vocoder allowing synthesis of high-fidelity speech with low computational complexity. StyleMelGAN employs temporal adaptive normalization to style a low-dimensional noise vector with the acoustic features of the target speech. For efficient training, multiple random-window discriminators adversarially evaluate the speech signal analyzed by a filter bank, with regularization provided by a multi-scale spectral reconstruction loss. The highly parallelizable speech generation is several times faster than real-time on CPUs and GPUs. MUSHRA and P.800 listening tests show that StyleMelGAN outperforms prior neural vocoders in copy-synthesis and Text-to-Speech scenarios.

Saad Alqithami, Musaad Alzahrani, Abdulkareem Alzahrani et al.

Attention Deficit Hyperactivity Disorder is one of the most common neurodevelopmental disorders in which patients have difficulties related to inattention, hyperactivity, and impulsivity. Those patients are in need of a psychological therapy use Cognitive Behavioral Therapy (CBT) to enhance the way they think and behave. This type of therapy is mostly common in treating patients with anxiety and depression but also is useful in treating autism, obsessive compulsive disorder and post-traumatic stress disorder. A major limitation of traditional CBT is that therapists may face difficulty in optimizing patients' neuropsychological stimulus following a specified treatment plan. Other limitations include availability, accessibility and level-of-experience of the therapists. Hence, this paper aims to design and simulate a generic cognitive model that can be used as an appropriate alternative treatment to traditional CBT, we term as "AR-Therapist." This model takes advantage of the current developments of augmented reality to engage patients in both real and virtual game-based environments.

Ahmed Mustafa, Arijit Biswas, Christian Bergler et al.

Classical parametric speech coding techniques provide a compact representation for speech signals. This affords a very low transmission rate but with a reduced perceptual quality of the reconstructed signals. Recently, autoregressive deep generative models such as WaveNet and SampleRNN have been used as speech vocoders to scale up the perceptual quality of the reconstructed signals without increasing the coding rate. However, such models suffer from a very slow signal generation mechanism due to their sample-by-sample modelling approach. In this work, we introduce a new methodology for neural speech vocoding based on generative adversarial networks (GANs). A fake speech signal is generated from a very compressed representation of the glottal excitation using conditional GANs as a deep generative model. This fake speech is then refined using the LPC parameters of the original speech signal to obtain a natural reconstruction. The reconstructed speech waveforms based on this approach show a higher perceptual quality than the classical vocoder counterparts according to subjective and objective evaluation scores for a dataset of 30 male and female speakers. Moreover, the usage of GANs enables to generate signals in one-shot compared to autoregressive generative models. This makes GANs promising for exploration to implement high-quality neural vocoders.