Advances in audio synthesis techniques have led to the creation of highly realistic audio deepfakes, posing growing threats to digital integrity and public trust. These synthetic manipulations mimic natural speech with high fidelity, making detection increasingly challenging and fueling the spread of misinformation, identity fraud, and voice-based attacks. To address these concerns, this study proposes the Adaptive Spectro-Temporal Diffusion Transformer (ASTDT), a novel detection framework that tackles key challenges in generalization, interpretability, and adaptability across diverse audio generation techniques. ASTDT integrates a score-based diffusion model to augment training spectrograms with realistic deepfake variations, improving generalization to unseen text-to-speech and voice conversion attacks. An adaptive spectro-temporal feature extraction mechanism partitions audio into interpretable frequency and temporal segments, while a dual-modal attention fusion module jointly processes magnitude and phase features. These fused features are processed by a transformer encoder with diffusion-aware attention, enabling effective modeling of long-range temporal dependencies. To enhance transparency, ASTDT includes an interpretability module that combines quantitative feature attributions and spatial heatmaps to explain model predictions. Experimental results across four benchmark datasets demonstrate the effectiveness of ASTDT, with the model achieving the lowest equal error rate of 1.20% on the ASVspoof 2019 dataset.

As advancements in agricultural technology unfold, machine learning and deep learning approaches are gaining interest in robust plant disease identification. Early disease detection, integral to agricultural productivity, has propelled innovations across all phases of detection. This survey paper provides a meticulous examination of plant disease detection systems, elucidating data collection methodologies and underscoring the pivotal role of datasets in model training. The narrative navigates through the complex areas of data and image processing techniques, segueing into an exploration of various segmentation methods. The survey emphasizes the importance of feature extraction and selection techniques, illustrating their efficacy in increasing classification accuracy. It examines the classification process, embracing both traditional machine learning and avant-garde deep learning methods, with a particular spotlight on Convolutional Neural Networks (CNNs). The study examines over one hundred seminal papers, anatomizing their dataset utilizations, feature considerations, and classification strategies. Overall, the paper contemplates the challenges permeating this vibrant field, addressing critical issues such as dataset diversity, model generalization, and real-world applicability.Note to Practitioners—To ensure crop health and yield, timely and precise plant disease detection is crucial. Our research, titled “Advances And Challenges in Plant Disease Detection: A Comprehensive Survey of Machine and Deep Learning Approaches”, examines the critical role of datasets, advanced image processing, and segmentation techniques in disease detection. This paper presents practitioners with a guide to the latest techniques for enhanced disease detection by emphasizing the significance of feature extraction and highlighting the capabilities of convolutional neural networks (CNNs). By understanding the highlighted challenges, such as dataset diversity and model generalization, industry professionals can better equip themselves to integrate these technological advancements into real-world agricultural applications.

Audio deepfake detection is emerging as a crucial field in digital media, as distinguishing real audio from deepfakes becomes increasingly challenging due to the advancement of deepfake technologies. These methods threaten information authenticity and pose serious security risks. Addressing this challenge, we propose a novel architecture that combines Convolutional Neural Networks (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) for effective deepfake audio detection. Our approach is distinguished by the feature concatenation of a comprehensive set of acoustic features: Mel Frequency Cepstral Coefficients (MFCC), Mel spectrograms, Constant Q Cepstral Coefficients (CQCC), and Constant-Q Transform (CQT) vectors. In the proposed architecture, features processed by a CNN are concatenated into two multi-dimensional features for comprehensive analysis, then analyzed by a BiLSTM network to capture temporal dynamics and contextual dependencies in audio data. This synergistic method ensures an understanding of both spatial and sequential audio characteristics. We validate our model on the ASVSpoof 2019 and FoR datasets, using accuracy and Equal Error Rate (EER) metrics for the evaluation.