Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM62323
Volume:
10 |
5 |
May 2026
Pages:
1-6
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.
department of computer science and engineering
shri shankaracharya college of engineering and technology
department of computer science and engineering
shri shankaracharya college of engineering and technology
department of computer science and engineering
shri shankaracharya college of engineering and technology
department of computer science and engineering
shri shankaracharya college of engineering and technology
AIR3D is a browser-based 3D modeling studio that leverages real-time hand gesture recognition via a standard webcam to enable fully touchless interaction with 3D content. Built on Three.js for WebGL rendering and MediaPipe Hands for 21-point hand landmark detection, AIR3D allows users to create 3D primitives such as boxes, spheres, cylinders, cones, tori, and prisms; manipulate objects through pinch-drag and two-hand scale or rotate gestures; draw freehand 3D lines and extrude them into surfaces; orbit and zoom the camera; and perform complete scene management including save and export. All interactions are driven purely by hand gestures captured through a standard webcam, with no plugins, no specialized hardware, and no installation required. This paper presents the system architecture including the layered z-index visual stack and shared APP state model; the gesture recognition pipeline; the Four-Pillar stabilization ap-proach comprising Exponential Moving Average smoothing, vote buffering, hysteretic pinch thresholds, and orbit iner-tia; the eight independent gesture state machines; mirror compensation logic; the dwell-click activation system; and the freehand 3D Draw Line system with extrusion capability. Results demonstrate that accurate, low-latency gesture control is achievable in a pure browser environment, with broad implications for accessibility, creative tools, education, and human-computer interaction research. The entire application runs from three plain files with no build step, no bundler, and no framework, demonstrating that powerful real-time AI-driven creative tools can be delivered as lightweight, de-pendency-free web applications.
KEYWORDS
AIR3D demonstrates that accurate, responsive, gesture-controlled 3D modeling is achievable in a standard web browser using only a webcam, without plugins, specialized hardware, or installation. The Four-Pillar stabilization approach (EMA smoothing, vote buffering, hysteretic thresholds, and orbit inertia) transforms noisy raw MediaPipe landmark data into stable gesture states suitable for creative 3D manipulation. Eight independent gesture state machines provide a rich vocabulary of natural interactions covering creation, manipulation, navigation, drawing, and scene management. The architecture of three plain files with no build step demonstrates that powerful real-time AI-driven creative tools can be delivered as accessible, dependency-free web applications. Future work includes multi-user collaboration, persistent cloud storage, export to standard 3D formats such as OBJ and GLB, voice command integration for hybrid multimodal input, and mobile device optimization for front-facing camera environments. AIR3D contributes a concrete existence proof that natural gesture interfaces for 3D content creation are viable in the browser today, with significant implications for accessibility, inclusive design, and the future of touchless creative tools.
[1] Leap Motion Inc., “Leap Motion Controller,” 2013. [Online]. Available: https://www.ultraleap.com [2] F. Zhang et al., “MediaPipe Hands: On-device Real-time Hand Tracking,” arXiv:2006.10214, 2020. [3] R. Cabello et al., “Three.js — JavaScript 3D Library,” GitHub. [Online]. Available: https://threejs.org [4] D. Kim, O. Hilliges, S. Izadi et al., “Digits: Freehand 3D Interactions Anywhere Using a Wrist-Worn Gloveless Sensor,” in Proc. ACM UIST, 2012, pp. 167–176. [5] A. Kuznetsov and I. Panchenko, “Smoothing and Stabilization Techniques for Real-Time Gesture Recognition Systems,” Journal of Computer Vision and Image Processing, vol. 11, no. 2, 2021. [6] Arunthathi S, Logeshwari T, and Anuratha V, “A Research Paper on Online Job Portal System,” International Journal of New Innovations in Engineering and Technology, vol. 24, no. 1, March 2024. (Reference format adapted for structural guidance.)
Publication Date
5/5/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
0
Total views since publication
PDF Downloads
5
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...