michiyomi

WALK DEMO

Walk through Ginza
for seven seconds.

This short demo is built from real michiyomi API data. It moves through a street from a pedestrian viewpoint and shows how the same location appears as structured, queryable language data.

The video has audio. Unmute the player to hear the short conversation; English captions are available from the captions menu.

Ginza · 6.7 seconds | Imagery adapted from Mapillary contributors (CC BY-SA 4.0) | In-video mini-map: GSI Tiles
English transcript

00:00–00:04 “Isn't the sidewalk around here a little narrow?”
00:04–00:05 “Yeah, it is.”

Original Japanese dialogue

「この辺りの歩道、ちょっと狭くない?」「そうね。」

How to read the screen

The left half is an instrument panel assembled from michiyomi API responses. The right is street imagery from a pedestrian viewpoint, with the current position at upper right.

LEFT PANEL 1

Load nearby contextdescribe_location

First it reports how much data exists near the coordinate—its coverage—then lists observations for the nearest scenes. Zero results means “not recorded,” not “nothing is there.”

LEFT PANEL 2

Evidence imageMapillary

The actual street image supporting the text. Each observation includes a scene ID and capture year so you can trace it back to its source image.

LEFT PANEL 3

Metadata — visual description

A description based only on visible content. Values such as “0.7 m walkable width” are monocular-image estimates and include confidence ratings.

LEFT PANEL 4

AI reasoning — what to say

The process of choosing one statement from the available evidence. Conditions are treated as belonging to the capture year and are not asserted as current facts.

UPPER RIGHT

Mini-map — current position

Shows the simulated walking position (map: GSI Tiles).

What runs behind the demo

STEP 1

Coordinate

The latitude and longitude of the walking position. That is the only input.

STEP 2

GET /v1/coverage

Reports the amount of available data before answering, so the user can see how much evidence exists.

STEP 3

GET /v1/scenes/nearby

Returns language data for nearby scenes, including each capture year.

The statements in the video are based on production /v1 API data. No registration or API key is required.

Hear this location in English

Fetch the production source record for the Ginza scene. This page presents and reads an English summary while preserving the API's authoritative Japanese observation below it.

GET /v1/scenes/1152690868598943
Press the button to fetch the production API record. Speech uses your device's English (en-US) voice.

Explore next