Phone-based transcription helps communicate spoken content, but it pulls attention away from the speaker. This concept uses augmented reality to keep captions, facial expressions, body language, and conversation context within the same field of view.

When captions appear on a phone or a separate screen, users must choose between reading the words and watching the person. That split attention makes it harder to read emotion, lip movements, and turn-taking, especially when several people are speaking.

I organized first-person frustrations and existing transcription limitations into communication, situational, visual, speaker, privacy, and real-time concerns. The synthesis highlighted the importance of speaker identification, readable customization, and preserving visual context.

Lip reading and context clues are central to many conversations, but mouth shapes vary and group discussions are difficult to track. The concept anchors text to the active speaker so users can connect language, facial expression, and speaker location more directly.

Placing captions directly below a speaker’s mouth reduces the distance between reading and lip observation. It also supports language learning by making it easier to connect the visible mouth shape with the spoken phrase and its translation.

The setup flow previews changes as users adjust text size, shape, color, line count, placement, scrolling, and hardware options. The goal is to let each person create a readable configuration independently before entering a live conversation.

Captions appear near the active speaker, while directional arrows indicate when someone is outside the user’s field of view. As the user turns, the text card resolves into position near that speaker, making group conversation easier to follow in real time.