UNICORN WHO DEV
← Back to catalogue
Repositories / Unicorn Who Dev

Vision Dataset Studio

An Android studio for batch image-dataset production with pre-annotation, human correction, controlled exports and optional on-device LiteRT learning.

Open the public source
Overview

An Android studio for batch image-dataset production with pre-annotation, human correction, controlled exports and optional on-device LiteRT learning. The public scope separates local dataset production, human validation, exports and the limits of on-device learning.

Repositories
1
Language
Kotlin
Activity
Sep 24, 26
Stars
1
Source
GitHub
Status
Connected source
Public dossier

Documentation and scope

Structured content from the public source documentation.

01

Overview

An Android studio for batch image-dataset production with pre-annotation, human correction, controlled exports and optional on-device LiteRT learning.

The public scope separates local dataset production, human validation, exports and the limits of on-device learning.

Visual referenceVision Dataset Studio — Android dataset tooling
02

Documented scope

  • Local or Hugging Face import
  • Pre-annotation followed by human correction and validation
  • Canonical, COCO, YOLO, WebDataset and vision-language exports
  • LiteRT learning disabled by default
  • Weights downloaded or imported after installation
  • Version 4.2.0-rc5 under qualification on September 23, 2026
03

Place in the catalogue

Vision Dataset Studio has a dedicated record connected to its current source.

04

Provenance and limits

The information is linked to the stated public source. Metrics reflect a reference snapshot and may change between synchronizations.

Provenance

A record traceable to its source

Editorial information, metadata and documentation remain connected to the original public repository.

Public identifier
Unicorn Who Dev/Vision Dataset Studio
Latest activity
September 24, 2026
Read the complete documentation
Project sources

Every source consolidated in this record

Public sources remain individually accessible. Private components are named without exposing their repository, revision or content.

01vision-dataset-studioGitHub
Technologies
androidkotlindatasetannotationliterton-device-learning