Action recognition (AR) has many applications, including surveillance, health/disabilities care, man-machine interactions, video-content-based monitoring, and activity recognition. Because human action videos contain a large number of frames, implemented models must minimize computation by reducing the number, size, and resolution of frames. We propose an improved method for detecting human actions in low-size and low-resolution videos by employing convolutional neural networks (CNNs) with channel attention mechanisms (CAMs) and autoencoders (AEs). By enhancing blocks with more representative features, convolutional layers extract discriminating features from various networks. Additionally, we use random sampling of frames before main processing to improve accuracy while employing less data. The goal is to increase performance while overcoming challenges such as overfitting, computational complexity, and uncertainty by utilizing CNN-CAM and AE. Identifying patterns and features associated with selective high-level performance is the next step. To validate the method, low-resolution and low-size video frames were used in the UCF50, UCF101, and HMDB51 datasets. Additionally, the algorithm has relatively minimal computational complexity. Consequently, the proposed method performs satisfactorily compared to other similar methods. It has accuracy estimates of 77.29, 98.87, and 97.16%, respectively, for HMDB51, UCF50, and UCF101 datasets. These results indicate that the method can effectively classify human actions. Furthermore, the proposed method can be used as a processing model for low-resolution and low-size video frames.

The high mortality rate and prevalence of cardiovascular disease (CVD) make early detection of the disease essential. Due to its simplicity and low cost, the phonocardiogram (PCG) system is widely used in healthcare applications for the recognition of CVD in multiclass problems. On the basis of the PCG signal, this paper proposes a hybrid method for classifying cardiac sounds with deep extracted features through two-step learning. For fine-grained features in Graph Convolutional Networks (GCNs), sampling and prior layers are employed. A PCG signal is divided into equal parts with overlap using the windowing process. L-spectrograms extract frequency-domain information from signals to figure out their power spectrum. Furthermore, the deep GCN tries to determine the association between CVD and spectrogram images to recognize CVD signals better. Combining retrieved features with convolutional neural network (CNN) characteristics reveals an image's intrinsic associations. To generate relational feature representations, correlations between clusters and GCN are visualized using a graph structure. CNN's discriminative ability has been enhanced by incorporating GCN attributes. Using Michigan Heart Sound and Murmur Database and PhysioNet/CinC 2016 Challenge results, we are 99.44% and 96.16% accurate, respectively. Through a combination of GCN architecture, CNN design, and deep features, the hybrid model significantly improves CVD classification accuracy. Measuring metrics demonstrate that the proposed approach detects CVD more effectively than previous approaches.