Perception: We believe in the future of autonomous technology to advance the quality of life for all humankind. We are pioneering AI research and development in this space to meet some of the toughest challenges in the safe deployment of autonomous vehicle road scenarios across the US, Europe and Vietnam.
Fueled with data from our own autonomous fleets, as well as from state-of-the-art simulation environments, our solutions are continuously and rapidly evolving. Our technologies are robust – backed by a world-class research team who tackle several fundamental and applied problems in autonomous technologies such as deep generative models, semi-supervised and unsupervised learning to scale – resulting in cost savings and the ability to rapidly scale up autonomous Perception across geographical locations and conditions.
Our Perception solution thrives on state-of-the-art deep learning, decision science, multi-sensor fusion technology and Lidar to deliver the best accuracy and performance in even the harshest environmental and driving scenarios. These include, but are not limited to: robust off-line and online sensor calibration; localization; lane detection; simulation and behavior cloning; 3D object detection; tracking; trajectory and behavior prediction; planning and control.
Voice, Language and
We are pushing the frontiers of assistant technology from: command and control to conversational AI on cars; edge/IoI devices to creating frictionless experiences in autonomous vehicles, smartphones and digital identity domains.
Automatic Speech Recognition (ASR)
We are committed to delivering the best ASR technology in both Vietnamese and English. Our current solution can accurately recognize Vietnam’s regional dialects with an almost instant response time. Most importantly, our solution is efficient, lightweight, off-line and does not require an Internet connection. It is fast, secure, privacy-preserving and ready to deploy not only in vehicles, but also in suitable IoT devices.
Text-to-Speech (TTS) and Dialogue Systems (DIA)
ASR, TTS and DIA complete the technology suite to innovate and deliver human-machine interaction products. We create the most human-friendly natural voices and run completely off-line on mobile devices.
An automobile’s carbon footprint and energy saving ability is critical. This is especially the case with electric vehicles to preserve battery efficiency. Our climate control technology is able to adjust in-car temperature and climate condition smartly with no input required from the driver. The technology learns from the driver's climate preferences and automatically adapts to the climate control function of the car, leaving a pleasant journey for the driver while, at the same time, saving battery usage.
We build technology to help improve user experience whether it is a driver, smartphone user, or an IoT consumer. Utilizing the user's past data on favorite songs, playing behaviors and context, our solution learns and recommends media items with the right content at the right time.
AI Edge Intelligence
We build technology for fast and accurate face detection and comparison, identification of dense facial landmarks and other facial attributes. Our face recognition solution ranks in the top 10 in NIST 2020, and is deployed to more than 1.6 million Vsmart phones.
Our technology precisely tracks & separates the speaker's voice from complex noise scenes to provide great suppression effects.
Our deep learning technology accurately recognizes thousands of common objects in photos and videos, fueling our perception and vision-based AI engines to serve our partners and customers.
AI Camera Under
A new trend in smartphone development is to have the camera hidden under the screen. In a world-first, we have successfully used AI to develop market-leading CUD technology to provide superior image quality. Deployed in millions of smartphones, our technology bypasses the optical barriers of the CUD, eliminating image blur, low light, color fidelity, noise and flare to create your favorite selfies with 4K high resolution.