MULTI-STYLE TRANSFORMATION APPARATUS AND METHOD

A device and method are designed to change the style of different parts of an image. First, the device divides the image into different sections based on the styles that need to be altered. Next, it creates a special mask for each section to apply the desired style. Then, a deep learning model processes the original image along with these masks. Finally, the result is an image that shows multiple styles applied to different areas.

ELECTRONIC DEVICE COMPRISING FLEXIBLE DISPLAY

An electronic device features a hinge that allows two parts, or housings, to fold together. It has a flexible display that is supported by both housings. This display includes a flat area and a bending section that curves around the edge of one housing. A buffer layer is placed between the display and the second housing to protect the bending part. Additionally, there are protective elements embedded in the buffer to provide extra safety for the display.

ARTIFICIAL INTELLIGENCE DEVICE FOR LIGHT TRANSFORMER-BASED EMOTION RECOGNITION (LTER) AND METHOD THEREOF

An artificial intelligence device can recognize emotions by analyzing both video and audio signals. It starts by breaking down the audio into a format called an audio embedding and the video into a visual embedding. Then, it processes these embeddings to create detailed features for both audio and visual information. By combining these features, the device can generate a comprehensive output. Finally, it uses this output to predict the emotion being expressed in the video and audio.

INTELLIGENT DOCUMENT CREATION AND REVIEW GENERATED BY A LARGE LANGUAGE MODEL

An intelligent system helps create and edit digital documents easily. Users can enter information into a special interface to generate or change documents. The system uses a large language model to create the document and shows it in the interface. It can also summarize the document and send that summary to another device for review. Additionally, the system can suggest changes to the document, allowing users to modify it based on those suggestions.

INTELLIGENT DOCUMENT CREATION AND REVIEW GENERATED BY A LARGE LANGUAGE MODEL

This technology helps people create and edit digital documents more easily. Users can enter information into a special interface, which then uses a large language model to generate a complete document. The system can also create a summary of the document for easy review by others. Additionally, it suggests changes to the document, allowing users to modify it based on these suggestions. Overall, this makes document creation and review faster and more efficient.

METHOD OF GENERATING FULLBODY ANIMATABLE PERSON AVATAR FROM SINGLE IMAGE OF PERSON, COMPUTING DEVICE AND COMPUTER-READABLE MEDIUM IMPLEMENTING THE SAME

A method allows the creation of a full-body animated avatar from just one image of a person. It starts by analyzing the image and using a model that understands the person's body shape and pose. Next, it defines how to apply textures to the avatar, linking each part of the body in the image to specific texture details. The system then samples colors and textures from the image to create a complete look for the avatar. Finally, this results in an animated character that can move and be customized based on the original photo.

DEVICE AND METHOD FOR NON-INVASIVE AND NON-CONTACT PHYSIOLOGICAL WELL BEING MONITORING AND VITAL SIGN ESTIMATION

A new health monitoring device can check a person's vital signs without needing to touch them. It uses a high-resolution camera to capture images of the user's face and applies advanced AI techniques to analyze these images. This analysis helps determine important health indicators like heart rate, respiratory rate, blood pressure, and oxygen levels. The device is designed to provide accurate results even in different conditions, with errors kept below 5%. It can be used in various areas such as telehealth, fitness tracking, public health checks, and safety in vehicles.

Method and System for Network of Generative AI Agents Representing Entities and Persons

A new system creates digital twins of people and entities using advanced AI agents. It collects various types of data from user devices, like text, audio, video, and images. This data is processed through special methods to build AI models that can understand and generate content in these formats. The models work together under user supervision and can adapt based on feedback from users. Once the models perform well enough, they can operate independently without direct user control.

GENERATING VIDEOS USING SEQUENCES OF GENERATIVE NEURAL NETWORKS

A new technology creates videos based on written descriptions. First, it takes a text prompt that describes a scene. Then, it uses a special program to understand the meaning of the text. After that, a series of advanced computer networks work together to turn this understanding into a video. The result is a final video that shows the scene described in the text.

VOICED-OVER MULTIMEDIA TRACK GENERATION

A method is created to change a video and audio track from one language to another. First, an audio model translates the original audio into the new language. Then, a video model adjusts the speaker's lip movements to match the new audio. After both tracks are ready, they are combined to create a complete video in the final language. This process helps make videos accessible to a wider audience by providing translations.