Artificial intelligence has evolved from mere tools to constant companions, fundamentally reshaping how we connect and communicate. As 2025 fast approaches, the fusion of AI and virtual environments accelerates at an unprecedented pace, and nowhere is this more evident than in the groundbreaking innovations to be showcased at the International Conference on Computer Graphics and Virtuality (ICCGV) and by visionary platforms like Soul App.
Since its inception in Tokyo in 2018, ICCGV, a leading global forum for AI and virtual interaction technologies, has become synonymous with innovation. This annual conference gathers top scholars, industry leaders, and innovators to present cutting-edge research, network, and discuss trends shaping the future. The 2025 conference, to be held in Chengdu, China, from February 21 to 23, promises to be an exciting confluence of ideas featuring topics like image processing, computational photography, and real-time rendering.
One of the highlights of ICCGV 2025 is the inclusion of Soul App’s pioneering paper, “LLM Gesticulator: Leveraging Large Language Models for Scalable and Controllable Co-Speech Gesture Synthesis.” Soul App’s paper demonstrates how large language models (LLMs) can generate synchronized, natural, full-body gestures that align with speech rhythms and semantic content. By combining text and audio inputs, the framework enables users to control gesture style and content, pushing the boundaries of virtual human interactions.
The paper’s significance lies not just in its technical innovation but in its implications for human-computer interaction. As online users increasingly seek emotional connections in their digital experiences, Soul App’s advancements provide an empathetic response to their needs. A survey by Soul App revealed that 60% of Gen Z users are open to AI-based friendships, appreciating the immediacy and lifelike qualities of AI interactions. Soul App’s technology directly responds to this demand, blending personalization and emotional resonance.
Since its infancy, Soul App has established itself as a leader in the AI-driven virtual companionship space. Its proprietary NAWA Engine, developed in 2020, combines AI, rendering, and image processing to create personalized 3D avatars and immersive social scenes. These innovations support real-time motion capture and restoration, enabling users to craft unique virtual personas. In 202, an in-house developed LLM for socializing, SoulX went live. It enables human-AI interactions filled with emotions and warmth, emerging as a key infrastructure for Soul to build AIGC-based social connections.The platform’s latest offerings are equally impressive. EchoVerse, a virtual environment introduced at the 2024 World Artificial Intelligence Conference, allows users to engage with versatile virtual characters in immersive real-time communications. Customizable through text descriptions and uploaded images, EchoVerse also lets users develop unique character voices. Another feature, Werewolf Awakening, integrates advanced AI reasoning into gaming scenarios, delivering dynamic, contextually appropriate interactions.
Soul App’s dedication to innovation extends to its research contributions. Its 2024 paper, “Multimodal Emotion Recognition with Vision-Language Prompting and Modality Dropout,” received recognition at the ACM Multimedia Conference. This study focused on improving emotion recognition accuracy through multimodal approaches, further highlighting Soul’s commitment to advancing AI capabilities in social applications.
The ICCGV 2025 conference offers a glimpse into the future of these technologies. Alongside Zhang Lu’s contributions through Soul App’s research, attendees will explore advancements in virtual reality, augmented reality, human-computer interaction, and immersive experiences. These developments enhance user engagement and redefine how technology bridges the gap between virtual and real-world interactions.
Soul App’s CTO, Tao Ming, encapsulates the company’s vision: “Soul is one of the most popular Internet platforms among Chinese young people. It is a natural application scene and a traffic entrance. Our advantage is that we can identify users’ actual needs and receive immediate feedback on the application of large models.” This user-centric approach and technological infrastructure position Soul App as a trailblazer in creating meaningful, personalized digital experiences.
Looking ahead, Soul App plans to further integrate 3D virtual human capabilities with multi-modal AI solutions. Zhang Lu aims to deliver even more natural, immersive, and intelligent interactive experiences, with profits being only a secondary goal. “Instead of seeking fortune from the metaverse, we think creating value is more important.”
As the platform expands its offerings, users will have unprecedented opportunities to showcase their personalities, connect emotionally, and explore the boundaries of human-computer interaction.
For those eager to delve deeper into this exciting intersection of AI and virtual reality, ICCGV 2025 is a must-attend event. It promises to showcase the latest innovations driving the future of immersive experiences. Whether through Soul App’s visionary technologies or the conference’s broader discussions, one thing is clear: the synergy between AI and human connection is only just beginning.
Zhang Lu uses Soul App to redefine virtual companionship, leading a movement transforming how we think about connection in the digital age. From customizable avatars and emotionally intelligent chatbots to cutting-edge gesture synthesis and immersive environments, the future of interaction is not just virtual—it’s profoundly personal. Whether you’re a tech enthusiast, an academic, or simply curious, now is the time to witness these groundbreaking changes and embrace a new dimension of human connection.
