Artificial Intelligence
Federated Learning (FL) as a promising distributed machine learning paradigm has been widely adopted in Artificial Intelligence of Things (AIoT) applications. However, the efficiency and inference capability of FL is seriously limited due to the presence of stragglers and data imbalance across massive AIoT devices, respectively. To address the above challenges, we present a novel asynchronous FL approach named CaBaFL, which includes a hierarchical \textbf{Ca}che-based aggregation mechanism and a feature \textbf{Ba}lance-guided device selection strategy.
- Categories:
56 Views
We introduce two novel datasets for cell motility and wound healing research: the Wound Healing Assay Dataset (WHAD) and the Cell Adhesion and Motility Assay Dataset (CAMAD). WHAD comprises time-lapse phase-contrast images of wound healing assays using genetically modified MCF10A and MCF7 cells, while CAMAD includes MDA-MB-231 and RAW264.7 cells cultured on various substrates. These datasets offer diverse experimental conditions, comprehensive annotations, and high-quality imaging data, addressing gaps in existing resources.
- Categories:
743 Views
This is a handwritten Chinese signatures dataset including offline images, online sequences and hand videos.
The TMS dataset consists of 50 writers, each contributing 10 multimodal genuine signatures and 10 people have additional 10 forgeries.
The dataset includes 60 subfiles (each for an individual writer) and a python file which processes the data. The name of forgery signature file starts with "F_". Each subfile includes 10 static images, 10 dynamic sequences and 10 videos.
(1) Images are in size of 1278×798 with RGBA channel, and stored in PNG format.
- Categories:
236 Views
The proposed SDTB dataset is collected from microscopic testicular tissue sections of 15 patients diagnosed with azoospermia. It simulates the process of selecting high-quality sperm in a testicular puncture scenario for further infertility diagnosis and treatment. Specifically, a testicular puncture is performed on these patients, followed by cleansing and shredding of the tubule tissue. The samples are then examined under a Nikon ECLIPSS Ti microscope at 200× magnification.
- Categories:
155 Views
The dataset has 99 rows, corresponding to 99 odor samples, their labels are shown in label.xlsx
The dataset has 450 columns, corresponding to the responses of 30 odor sensors under 15 heating voltages (2.6V-5.4V).
Taking the first row as an example, the first to 15th elements correspond to the response values of sensor No.1 at heating voltage range from 2.6 V to 5.4 V in 0.2 V increments; The 16th to 30th elements correspond to the response values of sensor No.2 at heating voltage range from 2.6 V to 5.4 V in 0.2 V increments; And so on.
- Categories:
170 Views
The decoupling method provides three key advantages compared to existing MRI reconstruction frameworks: it effectively incorporates prior knowledge of MRI structure using information from various angular perspectives, merges 4D data into multiple subspaces to streamline feature processing within each subspace, and showcases a flexible nature suitable for various tasks related to MR image processing.
- Categories:
23 Views
We evaluated the strategy performance on three different datasets (MNIST, FMNIST, and CIFAR10), which is simulated heterogeneity by assigning different data volume labels to these datasets. These datasets all consist of image data for vehicle perception tasks. The MNIST dataset contains 70,000 images from 10 different classes, including 60,000 train and 10,000 test samples . FMNIST and MNIST have similar data structures, both are grayscale images. In contrast, FMNIST focuses on more complex target recognition tasks, which contains 10 categories of everyday items .
- Categories:
73 Views
Six publicly available datasets are applied to Speech/Music Classification (SMC), Music Genre Classification (MGC), and Environmental Sound Classification (ESC), respectively. The utilized datasets include: 1) For SMC tasks, we employed the GTZAN-SMC and MUSAN datasets. 2) In ESC, the commonly employed ESC-10 and US8K datasets were included. 3) Classic GTZAN and Homburg datasets for MGC.
- Categories:
207 Views
The \textit{Plectropomus leopardus (P. leopardus)}, a species found in underwater environments, possesses substantial strategic importance due to its rich underwater resources. However, the natural habitat and industrial breeding environment of \textit{P. leopardus} is generally dark and complex, which presents notable challenges to object detection and recognition. In this research, we propose Plectropomus Leopardus recognition using Global Attention mechanism and Transfer learning(PLGAT), integrating a Global Attention Mechanism (GAM) with Transfer Learning to recognize \textit{P.
- Categories:
21 Views
Moroccan Dialect Emotion Recognition Dataset is a collection of voice records of people speaking Moroccan dialect in 5 states of emotion: Neutral, Happy, Sad, Angry and Fearful. The dataset has been collected in different Moroccan cities in 2024. Each recorder has 5 records per emotion class. The dataset contains 2000 record. The records are saved in .wav format, which is useful for signal processing with python libraries. The dataset is used for signal processing and emotion recognition using deep Learning models.
- Categories:
335 Views