About Me About Me About Me About Me About Me About Me About Me About Me About Me About Me
About Me About Me About Me About Me About Me About Me About Me About Me About Me About Me
I am a 4th year Computer Science and Cognitive Science undergraduate student at the University of Toronto. I enjoy all things visually creative, especially photography and graphic design, which are also my focus areas for developing interpretable, intuitive computational methods that assist the creation of such art.
I am grateful to have been able to work on a compact method for producing film-like colours from digital images, under the kind supervision of Prof. Michael Brown and Dr. Hakki Karaimer, at Samsung AI Center - Toronto, published as a poster for SIGGRAPH 2025.
Additionally, my hobbies inspire my work. See a selection of my photography, interactive media, and more, below.
Email: hyunjo.jang at mail.utoronto.ca
Research Research Research Research Research Research Research Research Research Research
Research Research Research Research Research Research Research Research Research Research
We present a compact, physically-based model that faithfully emulates the colour response of positive photographic film from a digital RAW image. Unlike hand-crafted look-up tables (LUT) or data-hungry neural networks, our approach analytically mirrors the film “capture-develop–scan” chain, all with around 30 trainable parameters. Least-squares optimisation is performed on 3168 colour patch pairs captured on one roll of Fujifilm VELVIA 100. Qualitative comparisons show the proposed model more closely matches real film than proprietary methods, and offers artefact-free rendering over discrete LUTs. The continuous model offers production-ready film emulation and a path for archival of discontinued stocks.
Photography Photography Photography Photography Photography Photography Photography
Photography Photography Photography Photography Photography Photography Photography
Interactive Media Interactive Media Interactive Media Interactive Media Interactive Media
Interactive Media Interactive Media Interactive Media Interactive Media Interactive Media
An interactive 3D scene of a train platform and an animated metro train, rendered in real-time in the Unity game engine. The visual theme was heavily inspired by the 1995 movie “Whisper of the Heart”, replicating the classic Japanese hand-drawn animation style. Selective, high contrast cel shading, edge outlines, and strong bloom were used as primary techniques. An animated train would arrive and depart upon user input or movement detection via an Xbox Kinect sensor. Part of an in-person interactive art exhibition in Gimpo, South Korea, in 2023. Made with Unity (main engine), Blender 3D (modelling), C# (scripting), and Xbox Kinect (person detection).
Born as an attempt to blur the lines between visually expressive but static posters and interactive game UI. Demonstrates an interface with a strong focus on stylised visual aesthetics while maintaining functionality. Also a compilation of various real-time 2D and 3D rendering techniques, including cel shading with outlines, character animation transitions, depth-based live shader replacement, and procedurally generated Bezier splines. Most visual effects were implemented in Unity’s URP Shadergraph. Made with Unity (main engine), VRoid Studio (character modelling), Blender 3D (model optimisation), C# (scripting).
Extras Extras Extras Extras Extras Extras Extras Extras Extras Extras Extras Extras
Extras Extras Extras Extras Extras Extras Extras Extras Extras Extras Extras Extras
