Neural Turtle Graphics for Modeling City Road Layouts

Abstract

Figure 1: We introduce Neural Turtle Graphics (NTG), a deep generative model for planar graphs. In the Figure, we show NTG's applications to (interactive) city road layout generation and parsing. AbstractWe propose Neural Turtle Graphics (NTG), a novel generative model for spatial graphs, and demonstrate its applications in modeling city road layouts. Specifically, we represent the road layout using a graph where nodes in the graph represent control points and edges in the graph represents road segments. NTG is a sequential generative model parameterized by a neural network. It iteratively generates a new node and an edge connecting to an existing node conditioned on the current graph. We train NTG on Open Street Map data and show that it outperforms existing approaches using a set of diverse performance metrics. Moreover, our method allows users to control styles of generated road layouts mimicking existing cities as well as to sketch parts of the city road layout to be synthesized. In addition to synthesis, the proposed NTG finds uses in an analytical task of aerial road parsing. Experimental results show that it achieves state-of-the-art performance on the SpaceNet dataset.

Semantic Labeling of Large-Area Geographic Regions Using Multiview and Multidate Satellite Images and Noisy OSM Training Labels

“…State-of-the-art approaches that demonstrate the use of labels derived from OSM for finding roads and/or buildings in overhead images include the studies described in [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20] and [21]. Many of these approaches use some category of neural networks as part of their machine-learning frameworks.…”

The Design and Optimization of Ship Cabin Space Layout Based on Crowd Simulation

“…为了追踪张量场中的流线, 从一 组种子点开始, 沿着流线生长的方向, 在迭代过程 中根据队列中的顶部元素生成流线, 同时将新种 子添加到队列中, 直到满足一定条件才停止. 空间布局建模工作可分为以建筑为基础的建 模 [28][29][30] 和以道路网络为基础的建模 [11,31]…”

Updating Street Maps using Changes Detected in Satellite Imagery

“…RoadTracer [4] and PolyMapper [18] propose an iterative tracing framework to extract road networks: they train a CNN to output the directionality of roads at each pixel, and employ an iterative search guided by the CNN to trace the road network. VecRoad extends the iterative tracing approach with a flexible step size and joint learning tasks [25], and Neural Turtle Graphics extends it with a sequential generative model [9]. Another recent technique, Sat2Graph, proposes a one-shot road extraction process where a CNN directly predicts the positions of road network vertices and edges [16].…”