Neural Graph Evolution: Towards Efficient Automatic Robot Design

Authors: Tingwu Wang, Yuhao Zhou, Sanja Fidler, Jimmy Ba

ICLR 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental As shown in experiments, NGE is the first algorithm that can automatically discover kinematically preferred robotic graph structures, such as a fish with two symmetric flat side-fins and a tail, or a cheetah with athletic front and back legs. In this section, we demonstrate the effectiveness of NGE on various evolution tasks.
Researcher Affiliation Collaboration 1 Department of Computer Science, University of Toronto 2 Vector Institute 3 NVIDIA
Pseudocode Yes Algorithm 1 Neural Graph Evolution
Open Source Code No The paper does not provide a statement or link to open-source code for the methodology described.
Open Datasets No Our experiments are simulated with Mu Jo Co. We design the following environments to test the algorithms. Fish Env: In the fish environment, graph consists of ellipsoids. The reward is the swimming-speed along the y-direction. We denote the reference human-engineered graph (Tassa et al., 2018) as GF . Walker Env: We also define a 2D environment walker constructed by cylinders, where the goal is to move along x-direction as fast as possible. No concrete access information for these custom environments.
Dataset Splits No The paper does not provide specific details on training, validation, or test dataset splits. It mentions training RL agents and evaluating fitness, but no explicit data partitioning information.
Hardware Specification Yes NGE is extremely efficient, it finds plausible robotic structures within a day on a single 64 CPU-core Amazon EC2 machine. A NGE session with 16-core m5.4xlarge ($0.768 per Hr) AWS machine can achieve almost the same performance with 64-core m4.16xlarge ($3.20 per Hr) in Fish environment in the same wall-clock time.
Software Dependencies No The paper mentions the use of 'Mu Jo Co' as a simulator but does not provide specific version numbers for any software dependencies.
Experiment Setup Yes We do a grid search on the hyper-parameters as summarized in Appendix E, and show the averaged curve of each method. Table 2: Hyperparameter grid search options. Table 3: Hyperparameters grid search options for NGE.