Search AI Products and News

Meta recently announced the global rollout of real-time translation for its Ray-Ban Meta smart glasses. Previously, this feature was limited to early testing users in select markets. This full launch allows users to enjoy more convenient language conversion across various scenarios, especially the ability to overcome language barriers offline. According to Meta, the real-time translation feature on Ray-Ban Meta smart glasses now covers global sales markets and supports English, French, and Italian (among other languages).

Meta has announced the rollout of several new features for its Ray-Ban smart glasses, including real-time translation, Instagram messaging, and calling. Initially available only to select users in a preview program, these features are now available to all Ray-Ban Stories users. The real-time translation feature, first revealed at Meta Connect 2024, underwent limited testing in select countries last December. Now, users can utilize this feature in supported markets.

In the rapidly evolving digital landscape, the convergence of virtual characters and interactive experiences is becoming increasingly prevalent. Persona Engine emerges as a comprehensive toolkit integrating Live2D, **Large Language Models (LLM), Automatic Speech Recognition (ASR), Text-to-Speech (TTS)**, and **Real-time Voice Cloning (RVC)**, offering users a seamless solution for creating vibrant and interactive digital characters. This engine is particularly well-suited for VTubing.

Google recently announced that the Gemini Live feature in its Gemini app is now available for free to all Android users. Initially, this feature was only accessible to Pixel 9 and Samsung Galaxy S25 users earlier this month through a Gemini Advanced subscription. However, due to positive user feedback, Google decided to expand access and make the feature widely available. The powerful Gemini Live functionality...

In the field of artificial intelligence, 3D vision and spatial understanding are becoming crucial for advancements in embodied AI, autonomous navigation, and virtual reality applications. In March 2025, Hangzhou Qunhe Technology announced the open-sourcing of its self-developed 3D vision large language model, SpatialLM, at the GTC2025 global conference, generating significant industry interest. This model, with its powerful spatial cognition capabilities and cost-effective data processing, offers revolutionary breakthroughs for robot training, architectural design, and AR/VR applications. AIbase is based on the latest information...

Microsoft recently unveiled a remarkable research project – WHAMM (World and Human Action MaskGIT Model). This innovative AI model can generate and run the classic game Quake II entirely within the AI model, producing a playable version in real-time. This research, part of Microsoft's Copilot Labs, explores the potential and boundaries of generative AI in interactive media. A revolutionary breakthrough.

Tencent Yuanbao announced the official launch of its latest V3 version, bringing significant feature upgrades. The core highlight of this update is the integration of the advanced HunYuan T1 and DeepSeek V3-0324 models. This significantly enhances Yuanbao's capabilities in code generation, structural understanding, and language response. Users only need to simply describe their needs.

RF-DETR is an open-source, real-time object detection model that allows for the detection of objects within images and videos. It is commercially usable, making it a versatile tool for various applications.

Google recently announced new capabilities for its Gemini Live, enabling it to answer user questions and "see" what's on a user's screen in real-time via a smartphone camera. These new features, developed under Google's Project Astra, are the culmination of nearly a year of research and development. According to Google spokesperson Alex Joseph, the Gemini Live updates mean users can expect a more intelligent and intuitive experience when using their phones with the assistant.

On March 13th, Sesame unveiled its latest speech synthesis model, CSM, attracting significant industry attention. According to the official introduction, CSM adopts an end-to-end Transformer-based multimodal learning architecture. It understands contextual information to generate natural and emotionally rich speech with stunningly realistic sound. The model supports real-time speech generation, processing both text and audio inputs. Users can also control features such as tone, intonation, rhythm, and emotion by adjusting parameters, showcasing high flexibility. CSM is considered a breakthrough in AI speech technology.

Once, AI 'vision' was limited, constrained by pre-defined labels. But the rules have changed! YOLOE, a groundbreaking AI model, has emerged, breaking free from the rigid constraints of traditional object detection. Imagine an AI that understands, not just through labeled data, but from text descriptions, blurry images, even with minimal context – a new era of universal real-time recognition!