Explore the vast range of titles available.

The accurate identification of crop pests and diseases is critical for global food security, yet the development of robust deep learning models is hindered by the limitations of existing datasets. To address this gap, we introduce DLCPD-25, a new large-scale, diverse, and publicly available benchmark dataset. We constructed DLCPD-25 by integrating 221,943 images from both online sources and extensive field collections, covering 23 crop types and 203 distinct classes of pests, diseases, and healthy states. A key feature of this dataset is its realistic complexity, including images from uncontrolled field environments and a natural long-tail class distribution, which contrasts with many existing datasets collected under controlled conditions. To validate its utility, we pre-trained several state-of-the-art self-supervised learning models (MAE, SimCLR v2, MoCo v3) on DLCPD-25. The learned representations, evaluated via linear probing, demonstrated strong performance, with the SimCLR v2 framework achieving a top accuracy of 72.1% and an F1 score (Macro F1) of 71.3% on a downstream classification task. Our results confirm that DLCPD-25 provides a valuable and challenging resource that can effectively support the training of generalizable models, paving the way for the development of comprehensive, real-world agricultural diagnostic systems.

Speech emotion recognition (SER) is one of the most exciting topics many researchers have recently been involved in. Although much research has been conducted recently on this topic, emotion recognition via non-verbal speech (known as the vocal burst) is still sparse. The vocal burst is concise and has meaningless content, which is harder to deal with than verbal speech. Therefore, in this paper, we proposed a self-relation attention and temporal awareness (SRA-TA) module to tackle this problem with vocal bursts, which could capture the dependency in a long-term period and focus on the salient parts of the audio signal as well. Our proposed method contains three main stages. Firstly, the latent features are extracted using a self-supervised learning model from the raw audio signal and its Mel-spectrogram. After the SRA-TA module is utilized to capture the valuable information from latent features, all features are concatenated and fed into ten individual fully-connected layers to predict the scores of 10 emotions. Our proposed method achieves a mean concordance correlation coefficient (CCC) of 0.7295 on the test set, which achieves the first ranking of the high-dimensional emotion task in the 2022 ACII Affective Vocal Burst Workshop & Challenge.

Starfish have a wide range of feeding habits, including starfish, sea urchins, sea cucumbers, corals, abalones, scallops, and many other marine organisms with economic or ecological value. The starfish outbreak in coastal areas will lead to severe economic losses in aquaculture and damage the ecological environment. However, the current monitoring methods are still artificial, time-consuming, and laborious. This study used an underwater observation platform with multiple sensors to observe the starfish outbreak in Weihai, Shandong Province. The platform could collect the temperature, salinity, depth, dissolved oxygen, conductivity, other water quality data, and underwater video data. Based on these data, the paper proposed an early warning model for starfish prevalence (EWSP) based on multi-sensor fusion. A deep learning-based object detection method extracts time-series information on the number of starfish from underwater video data. For the extracted starfish quantity information, the model uses the k-means clustering algorithm to divide the starfish prevalence level into four levels: no prevalence, mild prevalence, medium prevalence, and high prevalence. Correlation analysis concluded that the water quality factors most closely related to the starfish prevalence level are temperature and salinity. Therefore, the selected water quality factor and the number of historical starfish are inputted. The future starfish prevalence level of the starfish outbreak is used as an output to train the BP (back propagation) neural network to build EWSP based on multi-sensor fusion. Experiments show that the accuracy rate of this model is 97.26%, whose precision meets the needs of early warning for starfish outbreaks and has specific application feasibility.

Neonatal mortality remains a critical global challenge, particularly in resource-limited settings with restricted access to advanced diagnostic tools. Early detection of life-threatening conditions like Sepsis and Respiratory Distress Syndrome (RDS), which significantly contribute to neonatal deaths, is crucial for timely interventions and improved survival rates. This study investigates the use of newborn cry sounds, specifically the expiratory segments (the most informative parts of cry signals) as non-invasive biomarkers for early disease diagnosis. We utilized an expanded and balanced cry dataset, applying Self-Supervised Learning (SSL) models—wav2vec 2.0, WavLM, and HuBERT—to extract feature representations directly from raw cry audio signals. This eliminates the need for manual feature extraction while effectively capturing complex patterns associated with sepsis and RDS. A classifier consisting of a single fully connected layer was placed on top of the SSL models to classify newborns into Healthy, Sepsis, or RDS groups. We fine-tuned the SSL models and classifiers by optimizing hyperparameters using two learning rate strategies: linear and annealing. Results demonstrate that the annealing strategy consistently outperformed the linear strategy, with wav2vec 2.0 achieving the highest accuracy of approximately 90% (89.76%). These findings highlight the potential of integrating this method into Newborn Cry Diagnosis Systems (NCDSs). Such systems could assist medical staff in identifying critically ill newborns, prioritizing care, and improving neonatal outcomes through timely interventions.

Speech emotion recognition, as an important component of human-computer interaction technology, has received increasing attention. Recent studies have treated emotion recognition of speech signals as a multimodal task, due to its inclusion of the semantic features of two different modalities, i.e., audio and text. However, existing methods often fail in effectively represent features and capture correlations. This paper presents a multi-level circulant cross-modal Transformer (MLCCT) for multimodal speech emotion recognition. The proposed model can be divided into three steps, feature extraction, interaction and fusion. Self-supervised embedding models are introduced for feature extraction, which give a more powerful representation of the original data than those using spectrograms or audio features such as Mel-frequency cepstral coefficients (MFCCs) and low-level descriptors (LLDs). In particular, MLCCT contains two types of feature interaction processes, where a bidirectional Long Short-term Memory (Bi-LSTM) with circulant interaction mechanism is proposed for low-level features, while a two-stream residual cross-modal Transformer block is applied when high-level features are involved. Finally, we choose self-attention blocks for fusion and a fully connected layer to make predictions. To evaluate the performance of our proposed model, comprehensive experiments are conducted on three widely used benchmark datasets including IEMOCAP, MELD and CMU-MOSEI. The competitive results verify the effectiveness of our approach.

The potential of the sequence in Coronary Angiography (CA) frames for diagnosing coronary artery disease (CAD) has been largely overlooked. Our study aims to reveal the “Sequence Value” embedded within these frames and to explore methods for its application in diagnostics. We conduct a survey via Amazon Mturk (Mechanical Turk) to evaluate the effectiveness of Sequence Restoration Capability in indicating CAD. Furthermore, we develop a self-supervised deep learning model to automatically assess this capability. Additionally, we ensure the robustness of our results by differently selecting coronary angiographies/modules for statistical analysis. Our self-supervised deep learning model achieves an average AUC of 80.1% across five-fold validation, demonstrating robustness against static data noise and efficiency, with calculations completed within 30 s. This study uncovers significant insights into CAD diagnosis through the sequence value in coronary angiography. We successfully illustrate methodologies for harnessing this potential, contributing valuable knowledge to the field.

Political perspective detection in news media—identifying political bias in news articles—is an essential but challenging low-resource task. Prompt-based learning (i.e., discrete prompting and prompt tuning) achieves promising results in low-resource scenarios by adapting a pre-trained model to handle new tasks. However, these approaches suffer performance degradation when the target task involves a textual domain (e.g., a political domain) different from the pre-training task (e.g., masked language modeling on a general corpus). In this paper, we develop a novel multi-stage prompt tuning framework for political perspective detection. Our method involves two sequential stages: a domain- and task-specific prompt tuning stage. In the first stage, we tune the domain-specific prompts based on a masked political phrase prediction (MP3) task to adjust the language model to the political domain. In the second task-specific prompt tuning stage, we only tune task-specific prompts with a frozen language model and domain-specific prompts for downstream tasks. The experimental results demonstrate that our method significantly outperforms fine-tuning (i.e., model tuning) methods and state-of-the-art prompt tuning methods on the SemEval-2019 Task 4: Hyperpartisan News Detection and AllSides datasets.

We present a novel masked image modeling (MIM) approach, context autoencoder (CAE), for self-supervised representation pretraining. We pretrain an encoder by making predictions in the encoded representation space. The pretraining tasks include two tasks: masked representation prediction—predict the representations for the masked patches, and masked patch reconstruction—reconstruct the masked patches. The network is an encoder–regressor–decoder architecture: the encoder takes the visible patches as input; the regressor predicts the representations of the masked patches, which are expected to be aligned with the representations computed from the encoder, using the representations of visible patches and the positions of visible and masked patches; the decoder reconstructs the masked patches from the predicted encoded representations. The CAE design encourages the separation of learning the encoder (representation) from completing the pertaining tasks: masked representation prediction and masked patch reconstruction tasks, and making predictions in the encoded representation space empirically shows the benefit to representation learning. We demonstrate the effectiveness of our CAE through superior transfer performance in downstream tasks: semantic segmentation, object detection and instance segmentation, and classification. The code will be available at https://github.com/Atten4Vis/CAE.

Artificial intelligence (AI) is being increasingly integrated into scientific discovery to augment and accelerate research, helping scientists to generate hypotheses, design experiments, collect and interpret large datasets, and gain insights that might not have been possible using traditional scientific methods alone. Here we examine breakthroughs over the past decade that include self-supervised learning, which allows models to be trained on vast amounts of unlabelled data, and geometric deep learning, which leverages knowledge about the structure of scientific data to enhance model accuracy and efficiency. Generative AI methods can create designs, such as small-molecule drugs and proteins, by analysing diverse data modalities, including images and sequences. We discuss how these methods can help scientists throughout the scientific process and the central issues that remain despite such advances. Both developers and users of AI tools need a better understanding of when such approaches need improvement, and challenges posed by poor data quality and stewardship remain. These issues cut across scientific disciplines and require developing foundational algorithmic approaches that can contribute to scientific understanding or acquire it autonomously, making them critical areas of focus for AI innovation. The advances in artificial intelligence over the past decade are examined, with a discussion on how artificial intelligence systems can aid the scientific process and the central issues that remain despite advances.

Change detection (CD) in remote sensing (RS) imagery is a pivotal method for detecting changes in the Earth’s surface, finding wide applications in urban planning, disaster management, and national security. Recently, deep learning (DL) has experienced explosive growth and, with its superior capabilities in feature learning and pattern recognition, it has introduced innovative approaches to CD. This review explores the latest techniques, applications, and challenges in DL-based CD, examining them through the lens of various learning paradigms, including fully supervised, semi-supervised, weakly supervised, and unsupervised. Initially, the review introduces the basic network architectures for CD methods using DL. Then, it provides a comprehensive analysis of CD methods under different learning paradigms, summarizing commonly used frameworks. Additionally, an overview of publicly available datasets for CD is offered. Finally, the review addresses the opportunities and challenges in the field, including: (a) incomplete supervised CD, encompassing semi-supervised and weakly supervised methods, which is still in its infancy and requires further in-depth investigation; (b) the potential of self-supervised learning, offering significant opportunities for Few-shot and One-shot Learning of CD; (c) the development of Foundation Models, with their multi-task adaptability, providing new perspectives and tools for CD; and (d) the expansion of data sources, presenting both opportunities and challenges for multimodal CD. These areas suggest promising directions for future research in CD. In conclusion, this review aims to assist researchers in gaining a comprehensive understanding of the CD field.