BEIJING, June 6, 2023 /PRNewswire/ — WiMi Hologram Cloud Inc. (NASDAQ: WIMI) (“WiMi” or the “Company”), a leading global Hologram Augmented Reality (“AR”) Technology provider, today announced the development of a virtual interaction system based on gesture recognition. The system recognizes the user’s body and gesture movements through a recognition device. At the same time, the system uses immersive devices to provide feedback on the processing results, which significantly enriches the interaction semantics and enhances the natural friendliness of the interaction. It has the advantages of a realistic visualization effect, a friendly interaction mode, and a suitable compatibility interaction mode.

First, the system acquires real-time hand data through the acquisition device and determines the continuous points of hand gesture actions. Then the coarse-grained recognition of gesture actions is performed using a weighting-based dynamic time-regularization algorithm. After recognizing the basic gesture actions, the system extracts and matches the feature vectors to label the main local images. After that, the system performs fine-grained recognition of gesture actions using deep learning methods. The results of both are then combined to perform detailed gesture translation. Finally, the recognition results interact with the virtual environment to simulate an immersive feeling for the user.

The gesture recognition based on the weighted dynamic time regularization algorithm is divided into three main steps: building a gesture sample library, training the gesture sample model, and matching the similarity between the recognition sample and the sample library sample. In the recognition, some gesture frames are first intercepted as test samples and then compared with the samples in the sample library to find the most matching samples.

In the fine-grained gesture recognition process, the system automatically performs image feature extraction and selects appropriate gesture actions in candidate regions by convolutional neural networks. The algorithm includes a feature extraction network, region suggestion network acquisition, area of interest pooling, and fully connected layers. The gesture fine-grained image consists of several convolutional layers alternating with pooling layers for feature extraction. Each local information node of the convolutional network layer is connected to the local information node of the previous neural network layer through the convolutional network kernel. After optimization of the activation function, a feature map of the fine-grained gestures in this network layer is obtained. The input eigenface of each gesture in the pooled layer neural network corresponds to the eigenface information of the gesture in the previous layer. The system sparsely processes the feature information of the gestures to obtain features with spatially invariant rows. In the fully connected layer, the system integrates the feature information of gestures with considerable fine-grained differentiation. In the output layer, the system returns the classification processing results after the gesture fine-grained is feature-processed by the classifier.

In the feature extraction network, the system only focuses on the local details of the hand, such as finger positions, movements, and motion trajectories. After the feature extraction pooling operation, the dimensionality is reduced to half of the original image while ensuring the basic information of the image, which can effectively reduce the time complexity of the computation. The region suggests that the network takes the last output value of the feature extraction network as input and uses a fixed-size window to slide over the feature map. A vector value of low latitude is obtained with each slide, and each pixel point is returned to the location where the original image is located in the last layer of the region suggestion network.

The system inputs the extracted feature map and the candidate regions. Based on the input image, the system selects the region of interest, i.e., the location of the finger region corresponding to the feature map. The system performs a maximum pooling operation on the region and inputs this pooled data to the fully connected layer. Thus, the probability of belonging to the highest classification is input, while the offset of the position is obtained using border regression to obtain a more accurate target detection frame. This system can significantly improve the recognition rate of gesture actions and enhance the realistic experience and immersion of human-computer interaction.

Immersive virtual interaction based on gesture recognition is a hot topic in virtual reality, computer vision, and computer graphics. Along with the rapid development of technology and the gradual improvement of the human-computer interaction, people’s demand for visualization has also changed. The mouse, keyboard, and monitor can no longer give the user a three-dimensional experience. The two-dimensional interaction system alone makes certain interactive information challenging to express accurately. Therefore, WiMi’s system should have great application value.

About WIMI Hologram Cloud

WIMI Hologram Cloud, Inc. (NASDAQ:WIMI) is a holographic cloud comprehensive technical solution provider that focuses on professional areas including holographic AR automotive HUD software, 3D holographic pulse LiDAR, head-mounted light field holographic equipment, holographic semiconductor, holographic cloud software, holographic car navigation and others. Its services and holographic AR technologies include holographic AR automotive application, 3D holographic pulse LiDAR technology, holographic vision semiconductor technology, holographic software development, holographic AR advertising technology, holographic AR entertainment technology, holographic ARSDK payment, interactive holographic communication and other holographic AR technologies.

Safe Harbor Statements

This press release contains “forward-looking statements” within the Private Securities Litigation Reform Act of 1995. These forward-looking statements can be identified by terminology such as “will,” “expects,” “anticipates,” “future,” “intends,” “plans,” “believes,” “estimates,” and similar statements. Statements that are not historical facts, including statements about the Company’s beliefs and expectations, are forward-looking statements. Among other things, the business outlook and quotations from management in this press release and the Company’s strategic and operational plans contain forward−looking statements. The Company may also make written or oral forward−looking statements in its periodic reports to the US Securities and Exchange Commission (“SEC”) on Forms 20−F and 6−K, in its annual report to shareholders, in press releases, and other written materials, and in oral statements made by its officers, directors or employees to third parties. Forward-looking statements involve inherent risks and uncertainties. Several factors could cause actual results to differ materially from those contained in any forward−looking statement, including but not limited to the following: the Company’s goals and strategies; the Company’s future business development, financial condition, and results of operations; the expected growth of the AR holographic industry; and the Company’s expectations regarding demand for and market acceptance of its products and services.

Further information regarding these and other risks is included in the Company’s annual report on Form 20-F and the current report on Form 6-K and other documents filed with the SEC. All information provided in this press release is as of the date of this press release. The Company does not undertake any obligation to update any forward-looking statement except as required under applicable laws.