Artificial intelligence has come a long way from its beginning. In these 2 years, we’ve been expecting it to transit from a mere tool to a constant companion. Soul App is a prime example of this undeniable growth, demonstrating how technology can transform interpersonal connections.
The International Conference on Computer Graphics and Virtuality (ICCGV) is at the forefront of these dynamic intersections, serving as a global stage for innovators and visionaries.
ICCGV, which is hosted annually, features keynote lectures, presentations, and online sessions featuring human-computer interactions and technological innovation. It debuted its first conference in Tokyo in 2018 and has become a pivotal forum for cutting-edge research, scientific visualization, immersive experiences, and interactive graphics.
Conferences such as ICCGV have become crucial for tech industry collaboration. They provide opportunities for networking, knowledge exchange, and potential partnerships, offering access to the latest industry insights and trends.
The 2025 ICCGV conference will be held in Chengdu, China, from February 21 to 23. It will feature topics and discussions on image processing, computer vision, real-time rendering, and computational photography. Participants can submit papers for publication in the ICCGV 2025 Conference Proceedings by SPIE, which will be indexed by Ei Compendex, Scopus, and CPCI (Web of Science).
Soul App’s research contribution to next year’s ICCGV conference is the paper “LLM Gesticulator: Leveraging Large Language Models for Scalable and Controllable Co-Speech Gesture Synthesis.” The paper discusses a novel approach to virtual human interactions: using LLMs to generate synchronized and natural full-body gestures as a communication method.
The paper features multimodal integrations that combine text and audio inputs to produce gestures aligned with semantic content and speech rhythm. The framework demonstrates how scaling the LLM improves the quality of generated gestures for complex virtual interactions. Furthermore, it emphasizes how using text prompts allows users to control the content and style of each generated gesture.
In a previous Soul App survey, 60% of Gen Z tech users expressed being open to AI-based friendships. Why? They appreciate the immediate responses and find they are becoming more lifelike and less robotic—a problem Soul App seems to be solving well.
Soul App’s paper is just one of its many technological advancements. In addition to personalized virtual images and multimodal interaction solutions, Soul App has developed groundbreaking technologies, starting with the NAWA Engine in 2022. The self-developed engine combines AI, rendering, and image-processing capabilities to create personalized 3D avatars and immersive social scenes, supporting real-time motion capture and restoration.
A year later, Soul App launched its proprietary LLM, Soul X, which is capable of understanding nuanced context and engaging in emotionally resonant conversations. Both models have met younger generations’ needs for more meaningful and absorptive virtual interactions. They also prioritize user privacy, allowing individuals to express themselves more at ease.
Soul App’s ongoing research promises exciting changes in various sectors, including the gaming and virtual companionship industries. In 2024, Ubisoft launched neo-NPCs (neo non-playable characters) capable of generating real-time and contextually appropriate responses with greater situational awareness. Soul App has a massive opportunity to adopt this technology for more emergent storytelling opportunities and deeper connections.
“Soul is one of the most popular Internet platforms among Chinese youngsters. It is a natural application scene and a traffic entrance. Our advantage is we can identify the actual needs of users in the scene, and the application of the large model can also obtain feedback from users immediately,” says Tao Ming, Soul App’s CTO.
Soul App also unveiled three additional features at the 2024 World Artificial Intelligence Conference (WAIC). And the company’s features on display were Digital Twin, which allows users to create digital replicas of their personalities, greatly enhancing users’ socialising efficiency.
Werewolf Awakening is another case. The game interaction scenario features a showdown with AI entities possessing superior autonomous reasoning and speech from a combination of AI and real-life initiations.
Lastly, EchoVerse is a virtual space where users can engage in immersive real-time communications with versatile virtual characters and customize their personas through text descriptions and uploaded images. Users can also fuse and develop exclusive character voices based on the available timbres.
Nowadays, the multimodal interaction capabilities with Soul App are seemingly endless. According to Ming, “Users’ willingness to continuously engage in conversations and interactions with AI has already proved their recognition of the experience provided by the platform. It also indicates the sustainability of Soul’s insistence on the direction of “Synergic Model-application Interplay.”
As Soul App advances, users dabble deeper into the virtual companionship space to further personalize the social networking experience. Soul App is dedicated to creating more empathetic and enriching spaces for tech-savvy audiences to “naturalize” the blurred boundaries between real and virtual life.