98%
921
2 minutes
20
The pursuit of expressive and human-like music generation remains a significant challenge in the field of artificial intelligence (AI). While deep learning has advanced AI music composition and transcription, current models often struggle with long-term structural coherence and emotional nuance. This study presents a comparative analysis of three leading deep learning architectures: Long Short-Term Memory (LSTM) networks, Transformer models, and Generative Adversarial Networks (GANs), for AI-generated music composition and transcription using the MAESTRO dataset. Our key innovation lies in the integration of a dual evaluation framework that combines objective metrics (perplexity, harmonic consistency, and rhythmic entropy) with subjective human evaluations via a Mean Opinion Score (MOS) study involving 50 listeners. The Transformer model achieved the best overall performance (perplexity: 2.87, harmonic consistency: 79.4%, MOS: 4.3), indicating its superior ability to produce musically rich and expressive outputs. However, human compositions remained highest in perceptual quality (MOS: 4.8). Our findings provide a benchmarking foundation for future AI music systems and emphasize the need for emotion-aware modeling, real-time human-AI collaboration, and reinforcement learning to bridge the gap between machine-generated and human-performed music.
Download full-text PDF |
Source |
---|---|
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC12314053 | PMC |
http://dx.doi.org/10.1038/s41598-025-13064-6 | DOI Listing |
Hum Brain Mapp
September 2025
Tri-Institutional Center for Translational Research in Neuroimaging and Data Science (TReNDS), Georgia State University, Georgia Institute of Technology, and Emory University, Atlanta, Georgia, USA.
Investigating neuroimaging data to identify brain-based markers of mental illnesses has gained significant attention. Nevertheless, these endeavors encounter challenges arising from a reliance on symptoms and self-report assessments in making an initial diagnosis. The absence of biological data to delineate nosological categories hinders the provision of additional neurobiological insights into these disorders.
View Article and Find Full Text PDFComput Methods Biomech Biomed Engin
September 2025
Institute of Radio Physics and Electronics, University of Calcutta, Kolkata, India.
Parkinson's disease (PD) is a neurodegenerative condition that impairs motor functions. Accurate and early diagnosis is essential for enhancing well-being and ensuring effective treatment. This study proposes a deep learning-based approach for PD detection using EEG signals.
View Article and Find Full Text PDFEur J Case Rep Intern Med
August 2025
Internal Medicine, University of California, Riverside School of Medicine, Riverside, USA.
Introduction: Pulmonary embolism (PE) is a life-threatening condition with well-defined management strategies; however, the presence of a clot-in-transit (CIT)-a mobile thrombus within the right heart-introduces a uniquely high-risk scenario associated with a significantly elevated mortality rate. While several therapeutic approaches are available-including anticoagulation, systemic thrombolysis, surgical embolectomy, and catheter-directed therapies-there is no established consensus on a superior treatment modality. Catheter-based mechanical thrombectomy has emerged as a promising, minimally invasive alternative that mitigates the bleeding risks of systemic thrombolysis and the invasiveness of surgery.
View Article and Find Full Text PDFJ Clin Exp Hepatol
August 2025
Dept of Histopathology, PGIMER, Chandigarh, 160012, India.
Artificial intelligence (AI) is a technique or tool to simulate or emulate human "intelligence." Precision medicine or precision histology refers to the subpopulation-tailored diagnosis, therapeutics, and management of diseases with its sociocultural, behavioral, genomic, transcriptomic, and pharmaco-omic implications. The modern decade experiences a quantum leap in AI-based models in various aspects of daily routines including practice of precision medicine and histology.
View Article and Find Full Text PDFRadiol Adv
September 2024
Department of Radiology, Northwestern University and Northwestern Medicine, Chicago, IL, 60611, United States.
Background: In clinical practice, digital subtraction angiography (DSA) often suffers from misregistration artifact resulting from voluntary, respiratory, and cardiac motion during acquisition. Most prior efforts to register the background DSA mask to subsequent postcontrast images rely on key point registration using iterative optimization, which has limited real-time application.
Purpose: Leveraging state-of-the-art, unsupervised deep learning, we aim to develop a fast, deformable registration model to substantially reduce DSA misregistration in craniocervical angiography without compromising spatial resolution or introducing new artifacts.