The BeanPod can now display, listen, and speak, and it will even proactively remind you.

According to DeepTech Beat, ByteBeat has released SeedRealtime, which has been rolled out on the BeanPod App. It can process both audio and video simultaneously, allowing users to watch, listen, and respond at the same time. Compared to the previous generation Seeduplex, which could only listen and speak, this version has added real-time visual capabilities.

Users can use it to look at the camera to find things, check operations, or read text. When a target appears, the scene changes, or the user makes a mistake, it will proactively prompt without waiting for a question.

In a multitasking and noisy environment, it can also determine who is speaking, combining facial expressions, voice, and gestures to understand the context. Byte claims that issues such as interrupting, slow responses, and noise false positives have been reduced by about half compared to traditional stitching solutions.