Yunge Wen
Master’s of Science in Computer Science
@New York University
Machine Learning | Human-AI Interaction
yw3776@nyu.com
LinkedIn | Github | Animation Portfolio
← Try hovering over the portrait!
RESEARCH IN PROGRESS
Semantic Zooming Interface for Scalable Visualization of the Fly Neuron Connectome
The FlyWire dataset reconstructs full adult fly brain neurons using slicing and 3D reconstruction techniques, but large-scale neuronal networks often appear visually entangled, making the connectome hard to interpret. We are creating an interactive visualization tool that integrates morphological clustering and single-neuron downsampling with multi-resolution rendering and semantic zooming, enabling intuitive, Google Maps-style exploration and significantly reducing visual clutter.
Active Learning for Occlusion-Robust Sidewalk Segmentation
Tile2Net is an end-to-end tool that maps pedestrian infrastructure from aerial imagery via semantic segmentation, producing geo-referenced polygons and interconnected centerline networks. Building on this, we’re developing a batch labeling interface with active learning to improve occlusion handling and expand OpenStreetMap coverage.
Procedural Game Level Generation Through Story Graph
Traditional procedural content generation, based on rule-based or machine learning methods, often lacks the narrative depth required for story-driven games. Our approach leverages large language models to generate branching story arc graphs that drive dungeon generation, dynamically adjusting difficulty and visual assets to align with narrative progression.
"See What I Imagine, Imagine What I See": Human-AI Co-Creation System for 360° Panoramic Video Generation in VR
Imagine360 is a proof-of-concept prototype that advances panoramic video generation by integrating AI-driven co-creation, enabling speech-based prompts, perspective adjustments, and real-time VR customization, allowing users to seamlessly shape virtual environments through imagination.
Yunge Wen [arxiv]
PROJECTS
Bayesian Motion Trajectory Prediction
Finetuned YOLOv8 with the VisDrone dataset to enhance small object tracking, using Kalman filters to track single and multiple objects and predict their motion trajectories.
[Github]
Neural Style Transfer
Reproduced the 2015 seminal paper on image style transfer, with step-by-step visualization of content and style convolution results.
[Github]
Video2Video Search
Trained a convolutional autoencoder on Coco dataset. Extracted feature maps from video screenshots, stored in a vector database, and compared with query images through vector similarity.
[Github]
Enhancing LLM Accuracy with RAG
Created Huggingface WebApp to demonstrate retrieval-augmented generation for non-technical corporate users.
COMPUTATIONAL DESIGN
