Human-interaction-involved applications underscore the need for Multi-modal Sentiment Analysis (MSA). Although many approaches have been proposed to address the subtle emotions in different modalities, the power of explanations and temporal alignments is still underexplored. Thus, this paper proposes the Text-routed sparse mixture-of-Experts model with eXplanation and Temporal alignment for MSA (TEXT). TEXT first augments explanations for MSA via Multi-modal Large Language Models (MLLM), and then novelly aligns the representations of audio and video through a temporality-oriented neural network block. TEXT aligns different modalities with explanations and facilitates a new text-routed sparse mixture-of-experts with gate fusion. Our temporal alignment block merges the benefits of Mamba and temporal cross-attention. As a result, TEXT achieves the best performance across four datasets among all tested models, including three recently proposed approaches and three MLLMs. TEXT wins on at least four metrics out of all six metrics. For example, TEXT decreases the mean absolute error to 0.353 on the CH-SIMS dataset, which signifies a 13.5% decrement compared with recently proposed approaches.

Borescope inspection for aero-engine damage is a non-destructive endoscopic technique based on IoT devices and optical means, and serves as an important measure to ensure flight safety. However, borescope images of aero-engines present multiple challenges, including various damage types, large scale variations, numerous small targets. These issues pose significant obstacles to improving the accuracy of intelligent damage detection. Especially when the algorithm needs to be deployed on embedded devices for real-time processing, it is necessary to achieve coordinated optimization of detection accuracy and processing speed. To address the above problems, this paper proposes an efficient lightweight aero-engine damage detection model named GMA-Net for real-time detection on embedded devices. First, the C3Ghost module is introduced to construct the backbone network, which significantly reduces redundant computation while enhancing the model’s feature expression capability and receptive field. Second, the lightweight SegNext Attention module is embedded, which strengthens the model’s perception of tiny damage features through a combination of strip convolutions. Finally, a novel detection head structure is designed. By integrating the Detect MBConv module with lightweight depthwise separable convolution and the attention mechanism, it improves the efficiency of feature capture for multi-scale damage targets. Experiments on the self-built aero-engine damage dataset demonstrate that GMA-Net achieves promising results. Compared with the state-of-the-art lightweight detector YOLOv13n, GMA-Net reduces the number of parameters by 46.5% and lowers the computational cost (GFLOPs) by 37.1%. It also achieves a 0.86% improvement in mAP@50–95 and a 7.17% increase in recall. In addition, the model reaches 22.19 FPS on the Jetson Nano embedded device, meeting the efficiency requirements of practical intelligent aero-engine damage detection. The source code related to this paper is available at: https://github.com/Evking-silver/GMA-Net.

Predicting gaze point on mobile devices without calibration in unconstrained environments has great significance on human computer interaction. Appearance-based gaze estimation methods have been improved due to the recent advance in convolutional neural network (CNN) models and the availability of large-scale datasets. CNN models have limitations on extracting the global information of features and ignore the important information of local features. In this paper, we propose a novel structure named GazeAttentionNet. To improve the accuracy of gaze estimation, we use the global and local attention modules to utilize both global and local features. Firstly, we use MobileNetV2 and the self-attention layers as the global attention module to extract global features. Secondly, we add the local attention module containing the spatial attention to extract local features. With GazeAttentionNet, we achieve an excellent result on the GazeCapture dataset. The average errors of mobile phones and tablets are 1.67 cm and 2.37 cm.

Ambient Occlusion (AO) is a technique to approximate the effect of environment lighting and add realism to a scene by accentuating surface details and adding soft shadows, which is widely used in multimedia applications. Neural Network Ambient Occlusion (NNAO) is a pioneer in introducing deep learning to accurate and real-time AO, but it has two limitations: 1) performance bottleneck under excessive amount of samples ; 2) low contrast and blurred edges leading to unreal effect. To overcome these two limitations, we propose Accelerated Detail-enhanced Ambient Occlusion (ADAO) method based on NNAO by adopting three image processing methods: 1) spiral sampling in screen space; 2) contrast enhancement of AO map; 3) normal-depth edge preserving bilateral filtering. Experimental results show that the proposed method is over 2 times faster than NNAO and produces shadows with more realistic details.