REMY – Collecting Memories

REMY — an important part of the data collection process for participants in the REITION project. REMY will replace the first, simpler software application I developed several years ago.

But it’s also a local AI assistant that guides conversations, helps people retrieve memories, knowledge and emotions, and uses additional sensors to observe physiological responses.

The system combine an AI-led conversation with analysis of voice, facial expressions, gaze direction, pulse, skin conductance and four-channel EEG. This can help identify which people, places, photographs and events are especially meaningful to the user, allowing REMY to explore those memories through relevant follow-up questions.

REMY runs locally on the user’s computer, so private conversations, recordings and memories remains safe.

The project includes:

AI-guided conversation with audio and video recording

monitoring of EEG, pulse, head movement and skin conductance

analysis of facial expressions and gaze direction

recognition of people, objects and situations in photographs

creation of illustrations inspired by narrated memories

a memory map connecting people, places, events and periods of life

preservation of knowledge, beliefs, experiences and decision-making patterns

In the future,

REMY could support people experiencing memory loss by helping them recall important events, organise their life stories and communicate with their families.

The collected memories and knowledge could also form an interactive family or organisational archive, preserving a person’s stories, experiences, associations and way of thinking.

REMY is designed to help preserve some of the most valuable parts of a person’s life: memory, experience and personal history.

Additionally, the collected memories and recorded interactions may serve as the foundation for creating simple family or professional digital avatars that preserve a person’s conversational style, memories, and accumulated knowledge.

For professional applications, these avatars could be further enhanced with more advanced decision-support capabilities based on real-world experience, individual reasoning patterns, and the cognitive processes characteristic of a particular person.