Video watermarking is an important means of video and multimedia copyright protection, but the current watermarking algorithm is difficult to ensure high robustness under various attacks. In this paper, a video watermarking algorithm based on NSCT, pseudo 3D-DCT and NMF has been proposed. Combined with NSCT, 3D-DCT and NMF, the algorithm embeds the encrypted QR code copyright watermark into the NMF base matrix to improve the anti-attack ability of the watermark under the condition of invisibility. The experimental results show that the algorithm ensures the invisibility of the watermark with a high signal-to-noise ratio of the video, and meanwhile has high ability and robustness against common single and combined attacks, such as filtering, noise, compression, shear, rotation and so on. The issue that the video watermarking algorithm has poor resistance to various attacks, especially the shearing attack, has been solved in this paper; thus, it can be used for digital multimedia video copyright protection.

In sign language video, the hand region is small, the resolution is low, the motion speed is fast, and there are cross occlusion and blur phenomena, which have a great impact on sign language recognition rate and speed, and are important factors restricting sign language recognition performance. To solve these problems, this paper proposes an improved 3D-ResNet sign language recognition algorithm with enhanced hand features, aiming to highlight the features of both hands, solve the problem of missing more effective information when relying only on global features, and improve the accuracy of sign language recognition. The proposed method has two improvements. Firstly, the algorithm detects the left and right hand regions based on the improved EfficientDet network, uses the improved Bi-FPN module and dual channel and spatial attention module are used to enhance the detection ability of the network for small targets like hand. Secondly, the improved residual module is used to improve the 3D-ResNet18 network to extract sign language features. The global, the left-hand and the right-hand image sequences are divided into three branches for feature extraction and fusion, so as to strengthen the attention to hand features, strengthen the representation ability of sign language features, and achieve the purpose of improving the accuracy of sign language recognition. In order to verify the performance of this algorithm, a series of experiments are carried out on CSL dataset. For example, in the experiments of hand detection algorithm and sign language recognition algorithm, the performance indicators such as Top-N, mAP, FLOPs and Parm are applied to find the optimal algorithm framework. The experimental results show that the Top1 recognition accuracy of this algorithm reaches 91.12%, which is more than 10% higher than that of C3D, P3D and 3D-ResNet basic networks. From the performance indicators of Top-N, mAP, FLOPs, Parm and so on, the performance of the algorithm in this paper is better than several algorithms in recent three years, such as I3D+BLSTM, B3D ResNet, AM-ResC3D+RCNN and so on. The results show that the hand detection network with enhanced hand features and three-dimensional convolutional neural network proposed in this paper can achieve higher accuracy of sign language recognition.

Human action recognition is an important part for computers to understand the behavior of people in pictures or videos. In a single image, there is no context information for recognition, so its accuracy still needs to be greatly improved. In this paper, a single-image human action recognition method based on improved ResNet and skeletal keypoints is proposed, and the accuracy is improved by several methods. We improved the backbone network ResNet-50 and CPN to a certain extent and constructed a multitask network to suit the human action recognition task, which not only improves the accuracy but also balances the total number of parameters and solves the problem of large network and slow operation. In this paper, the improvement methods of ResNet-50, CPN, and whole network are tested, respectively. The results show that the single-image human action recognition based on improved ResNet and skeletal keypoints can accurately identify human action in the case of different human movements, different background light, and occlusion. Compared with the original network and the main human action recognition algorithms, the accuracy of our method has its certain advantages.

In response to the current challenges fire detection algorithms encounter, including low detection accuracy and limited recognition rates for small fire targets in complex environments, we present a lightweight fire detection algorithm based on an improved YOLOv5s. The introduction of the CoT (Contextual Transformer) structure into the backbone neural network, along with the creation of the novel CSP1_CoT (Cross stage partial 1_contextual transformer) module, has effectively reduced the model’s parameter count while simultaneously enhancing the feature extraction and fusion capabilities of the backbone network; The network’s Neck architecture has been extended by introducing a dedicated detection layer tailored for small targets and incorporating the SE (Squeeze-and-Excitation) attention mechanism. This augmentation, while minimizing parameter proliferation, has significantly bolstered the interaction of multi-feature information, resulting in an enhanced small target detection capability; The substitution of the original loss function with the Focal-EIoU (Focal-Efficient IoU) loss function has yielded a further improvement in the model’s convergence speed and precision; The experimental results indicate that the modified model achieves an mAP@.5 of 96% and an accuracy of 94.8%, marking improvements of 8.8% and 8.9%, respectively, over the original model. Furthermore, the model’s parameter count has been reduced by 1.1%, resulting in a compact model size of only 14.6MB. Additionally, the detection speed has reached 85 FPS (Frames Per Second), thus satisfying real-time detection requirements. This enhancement in precision and accuracy, while simultaneously meeting real-time and lightweight constraints, effectively caters to the demands of fire detection.