Qualcomm presented this Thursday its new technology solution for video calling and video conferencing called Qualcomm QCS8250 Video Collaboration Reference Design, a platform developed for unified communications applications with on-device AI. It also improves audio and video quality. As we wanted to demonstrate, technology companies are all throwing themselves on intelligence artificial. But let's see the details of Qualcomm QCS8250 Video Collaboration Reference Design.
Qualcomm wants to improve video conferencing, virtual meetings and collaboration in modern hybrid workplaces with AI
According to the company, this innovation was launched with the aim of "improving virtual meetings and collaboration in today's modern hybrid workplace" by eliminating distractions and increasing productivity during Internet meetings by offering a more engaging. Dev Singh, vice president of Qualcomm, said:
Qualcomm continues to push the boundaries of AI in audio, camera and video, and our latest benchmark Qualcomm QCS8250 enables more immersive, near-seamless video conferencing experiences that ensure all colleagues, in person or remotely, can be clearly seen and heard
Read also: Netflix account sharing 2023: what has changed since March
There are three essential points of this new AI-based solution:
- AI Group Framing: Video AI will automatically detect people in the conference room and seamlessly zoom in or out using its three Ultra HD cameras (ultra-wide, wide-angle and telephoto) to provide an optimal meeting view of all conference participants;
- AI Speaker Framing: by combining Qualcomm audio capture with video AI and virtually switching between three 4K UltraHD cameras, it can identify a person who is speaking and zoom in on that person to provide a more detailed and accurate view close to the presenter. An auto-focus in essence;
- AI noise suppression: state-of-the-art neural network-based noise canceling technology isolates the speaker's voice and eliminates distracting background noises such as food packaging opening, paper moving, pen clicking or keyboard typing , without the need for multiple microphones required by traditional solutions.
