Communication for hearing-impaired communities is an exceedingly challenging task, which is why dynamic sign language was developed. Hand gestures and body movements are used to represent vocabulary in dynamic sign language. However, dynamic sign language faces some challenges, such as recognizing complicated hand gestures and low recognition accuracy, in addition to each vocabulary’s reliance on a series of frames. This paper used MediaPipe in conjunction with RNN models to address dynamic sign language recognition issues. MediaPipe was used to determine the location, shape, and orientation by extracting keypoints of the hands, body, and face. RNN models such as GRU, LSTM, and Bi-directional LSTM address the issue of frame dependency in sign movement. Due to the lack of video-based datasets for sign language, the DSL10-Dataset was created. DSL10-Dataset contains ten vocabularies that were repeated 75 times by five signers providing the guiding steps for creating such one. Two experiments are carried out on our dataset (DSL10-Dataset) using RNN models to compare the accuracy of dynamic sign language recognition with and without the use of face keypoints. Experiments revealed that our model had an accuracy of more than 99%.

Early and accurate detection of breast cancer is essential for reducing mortality rates and improving clinical outcomes. However, deep learning (DL) models used in healthcare face significant challenges, including concerns about data privacy, domain-specific overfitting, and limited interpretability. To address these issues, we propose BreastSwinFedNetX, a federated learning (FL)-enabled ensemble system that combines four hierarchical variants of the Swin Transformer (Tiny, Small, Base, and Large) with a Random Forest (RF) meta-learner. By utilizing FL, our approach ensures collaborative model training across decentralized and institution-specific datasets while preserving data locality and preventing raw patient data exposure. The model exhibits strong generalization and performs exceptionally well across five benchmark datasets—BreakHis, BUSI, INbreast, CBIS-DDSM, and a Combined dataset—achieving an F1 score of 99.34% on BreakHis, a PR AUC of 98.89% on INbreast, and a Matthews Correlation Coefficient (MCC) of 99.61% on the Combined dataset. To enhance transparency and clinical adoption, we incorporate explainable AI (XAI) through Grad-CAM, which highlights class-discriminative features. Additionally, we deploy the model in a real-time web application that supports uncertainty-aware predictions and clinician interaction and ensures compliance with GDPR and HIPAA through secure federated deployment. Extensive ablation studies and paired statistical analyses further confirm the significance and robustness of each architectural component. By integrating transformer-based architectures, secure collaborative training, and explainable outputs, BreastSwinFedNetX provides a scalable and trustworthy AI solution for real-world breast cancer diagnostics.

In the discipline of hand gesture and dynamic sign language recognition, deep learning approaches with high computational complexity and a wide range of parameters have been an extremely remarkable success. However, the implementation of sign language recognition applications for mobile phones with restricted storage and computing capacities is usually greatly constrained by those limited resources. In light of this situation, we suggest lightweight deep neural networks with advanced processing for real-time dynamic sign language recognition (DSLR). This paper presents a DSLR application to minimize the gap between hearing-impaired communities and regular society. The DSLR application was developed using two robust deep learning models, the GRU and the 1D CNN, combined with the MediaPipe framework. In this paper, the authors implement advanced processes to solve most of the DSLR problems, especially in real-time detection, e.g., differences in depth and location. The solution method consists of three main parts. First, the input dataset is preprocessed with our algorithm to standardize the number of frames. Then, the MediaPipe framework extracts hands and poses landmarks (features) to detect and locate them. Finally, the features of the models are passed after processing the unification of the depth and location of the body to recognize the DSL accurately. To accomplish this, the authors built a new American video-based sign dataset and named it DSL-46. DSL-46 contains 46 daily used signs that were presented with all the needed details and properties for recording the new dataset. The results of the experiments show that the presented solution method can recognize dynamic signs extremely fast and accurately, even in real-time detection. The DSLR reaches an accuracy of 98.8%, 99.84%, and 88.40% on the DSL-46, LSA64, and LIBRAS-BSL datasets, respectively.

Meticulous learning of human emotions through speech is an indispensable function of modern speech emotion recognition (SER) models. Consequently, deriving and interpreting various crucial speech features from raw speech data are complicated responsibilities in terms of modeling to improve performance. Therefore, in this study, we developed a novel SER model via attention-oriented parallel convolutional neural network (CNN) encoders that parallelly acquire important features that are used for emotion classification. Particularly, MFCC, paralinguistic, and speech spectrogram features were derived and encoded by designing different CNN architectures individually for the features, and the encoded features were fed to attention mechanisms for further representation, and then classified. Empirical veracity executed on EMO-DB and IEMOCAP open datasets, and the results showed that the proposed model is more efficient than the baseline models. Especially, weighted accuracy (WA) and unweighted accuracy (UA) of the proposed model were equal to 71.8% and 70.9% in EMO-DB dataset scenario, respectively. Moreover, WA and UA rates were 72.4% and 71.1% with the IEMOCAP dataset.

Today's overpopulation and fast urbanization present a significant challenge for developing countries in the form of excessive garbage generation. Managing waste is essential in creating sustainable and habitable communities, but it remains an issue for developing countries. Finding an efficient smart waste management system is a challenge in current research. In recent years, robots and artificial intelligence have influenced a wide range of industries, especially waste management. This research proposes a waste segregation system that integrates the robot arm and YOLOv6 object detection model to automatically sort the garbage according to its type and achieve real-time requirements. The proposed algorithm utilizes the pros of the hardware-friendly architecture of YOLOv6 while keeping high detection accuracy in detecting and classifying garbage. Moreover, the proposed system creates a 3D model of a 4 DOF robotic arm by CAD tools. A new approach based on a geometric method is proposed to solve the inverse kinematics problem in order to precisely calculate the proper angles of the robot arm's joints via a unique solution with less computational time. The proposed system is evaluated on a modified TrashNet dataset with seven garbage classes. The experiments reveal that the proposed algorithm outperforms the other recent YOLO models in terms of precision, recall, F1 score, and model size. Furthermore, the proposed algorithm consumes approximately fractions of a second for picking up and placing a single object in its proper basket.