Yutong Liang | 梁昱桐

Hi there! I am Yutong Liang, a senior undergraduate student at the School of EECS , Peking University, majoring in Computer Science.

I'm interested in the field of character animation/embodied AI, supervised by Prof. Libin Liu in the VCL lab of Peking University.

Currently my interest is in HOI (human-object interaction) , HSI (human-scene interaction) and hands!

Previously I researched on the field of Data Structures in Network Measurement, supervised by Prof. Tong Yang.

I was a research intern at Stanford University during the summer of 2024, worked with Michelle Guo and Pei Xu in the TML Lab and SVL Lab, under the guidance of Prof. Karen Liu and Prof. Jiajun Wu.

Click the image to see new photos!

To facilitate daily activities and movement, indoor layouts must be easy to navigate and comfortable. Therefore, we integrate motion considerations into the design process. For one particular scene, we used motion synthesis and inverse dynamics to evaluate it, followed by optimization with CMA-ES. This approach significantly improved metrics related to human movement distance and energy consumption for grabbing objects. In the future, we aim to refine our metric calculations and incorporate more personalized characteristics into indoor scene design.

Slides [Google Slides]

We propose SimiSketch, a sketching algorithm designed to tackle this specific problem of estimating similarity between sets. We first present two simpler versions that employ intuitive sketches for similarity estimation. Subsequently, we formally introduce SimiSketch and leverage SALSA to enhance accuracy. To validate our algorithms, we conduct extensive testing on synthetic datasets, real-world network traffic, and text articles. Our experiment shows that compared with the state-of-the-art, SimiSketch can improve the accuracy by up to 42 times, and increase the throughput by up to 360 times. The complete source code is open-sourced and available on GitHub for reference.

Paper [Arxiv] | Code [GitHub]