Title:Physically Based Rendering
Speaker: Ji Li
PBR (Physically-Based Rendering). Generally speaking, it can be seen from the literal meaning that this is a rendering technology based on the simulation of physical laws. It was first used for photorealistic rendering of movies. In recent years, due to the continuous improvement of hardware performance, it has been widely used in real-time rendering of PC games andconsole games. PBR contains a lot of knowledge, includes Diffuse and reflection Conservation of energy Microfacet material , etc. In this seminar, we will Popularize PBR and its related technologies.
Supervisor: Xiao Liang
Title:HOW POWERFUL ARE GRAPH NEURAL NETWORKS?
Speaker: Jincheng Huang
Graph Neural Networks (GNNs) are an effective framework for representation learning of graphs. GNNs follow a neighborhood aggregation scheme, where the representation vector of a node is computed by recursively aggregating and transforming representation vectors of its neighboring nodes. Many GNN variants have been proposed and have achieved state-of-the-art results on both node and graph classification tasks. However, despite GNNs revolutionizing graph representation learning, there is limited understanding of their representational properties and limitations. Here, we present a theoretical framework for analyzing the expressive power of GNNs to capture different graph structures. Our results characterize the discriminative power of popular GNN variants, such as Graph Convolutional Networks and GraphSAGE, and show that they cannot learn to distinguish certain simple graph structures. We then develop a simple architecture that is provably the most expressive among the class of GNNs and is as powerful as the WeisfeilerLehman graph isomorphism test. We empirically validate our theoretical findings on a number of graph classification benchmarks, and demonstrate that our model achieves state-of-the-art performance.
Supervisor: Biao Wang, Yang Yang