Skip to content
All work
P-11activeInteractive Installation · 2026

NeiseDump! — Local Web Photobooth

A local web photobooth with stylized live capture, OpenCV-based face labeling, and a fixed 9:16 collage output pipeline for event-ready image composition.

Role
Developer · Web Capture & Visual Pipeline
Timeframe
Mar 2026 - Present
Domain
Interactive Installation
Status
active
Layout
9:16 Collage
Preview
4:3 Camera Feed
Backend
Local OpenCV
NeiseDump! — Local Web Photobooth

Overview

NeiseDump! is a local-first photobooth app that serves a stylized camera interface through a web frontend and sends face detection to a local backend.

The final output is a vertical collage sheet with a fixed layout, designed for clean event-style visual exports rather than generic webcam dumping.

The Challenge

The main job was to keep capture, face labeling, theme switching, and collage generation consistent across the browser and the local backend without making the system fragile.

Approach

  • 01Served the frontend as a local web app with a lightweight static-server workflow.
  • 02Used a local OpenCV backend for face detection and label overlay.
  • 03Applied a hard two-color duotone style to the live feed and output result.
  • 04Built a fixed 9:16 composition with header, four photo slots, and footer branding.

Outcomes

  • Delivered a stylized photobooth with a clear local-only capture pipeline.
  • Produced a consistent collage output suitable for event display or sharing.
  • Kept the implementation simple enough to run from a local development workflow.

Stack

Frontend

  • JavaScript
  • HTML
  • CSS

Backend

  • Python
  • OpenCV
  • Local detection server

Output

  • 9:16 collage
  • Duotone processing
  • Static asset pipeline

Interested in a build like this?

I take on instrumentation, embedded, and computational physics work. Let's talk about your measurement problem.

Gilang Pratama Putra Siswanto