Ben (Simiao) Ren, Ph.D.

Ben (Simiao) Ren, Ph.D.

Co-Founder at ScamAI | Ex-Meta AI | ML Ph.D. from Duke

First (co) author publications

Benchmarking Data-driven Surrogate Simulators for Artificial Electromagnetic Materials

Published in **NeurIPS 2021**: Advances in Neural Information Processing Systems 34, 2021

In press. Co-first author paper about creating a benchmark suite for easy benchmarking of the forward mapping of the AEM property emulator

Recommended citation: Deng, Y., Dong, J., Ren, S., Khatib, O., Soltani, M., Tarokh, V., ... & Malof, J. (2021). Benchmarking Data-driven Surrogate Simulators for Artificial Electromagnetic Materials. https://openreview.net/forum?id=-or413Lh_aF

Benchmarking deep inverse models over time, and the neural-adjoint method

Published in **NeurIPS 2020**: Advances in Neural Information Processing Systems 33, 2020

This paper is about deep learning and inverse problems (scientific discovery but not imagery). We benchmarked state-of-the-art deep inverse models and proposed a new algorithm called neural-adjoint that accurately and efficiently solves the inverse problems

Recommended citation: Ren, S., Padilla, W., & Malof, J. (2020). Benchmarking Deep Inverse Models over time, and the Neural-Adjoint method. In H. Larochelle, M. Ranzato, R. Hadsell, M. F. Balcan, & H. Lin (Eds.), Advances in Neural Information Processing Systems (Vol. 33, pp. 38–48). Curran Associates, Inc. https://proceedings.neurips.cc/paper/2020/hash/007ff380ee5ac49ffc34442f5c2a2b86-Abstract.html

A modular view of recent deep inverse modelsfor inverse design problems

Published in Submitting, 1997

Submitting. In this work on provide a modular view for the recent deep inverse models for inverse design problem and tested their performance when we mix-n-match their individual components together

Recommended citation: Ren, S., Padilla, W., Malof, J., (2021). A modular view of recent deep inverse modelsfor inverse design problems.

Inverse deep learning methods and benchmarks for artificial electromagnetic material design

Published in Submitting, 1997

Submitting. First author paper focusing on benchmarking the various deep learning models in specifically the realm of meta-material design.

Recommended citation: Ren, S., Mahendra, A., Deng, Y., Khatib, O., Padilla, W. J. & Malof, J. (2021). Inverse deep learning methods and benchmarks for artificial electromagnetic material design https://onlinelibrary.wiley.com/doi/full/10.1002/adfm.202101748

Mapping solar photovoltaic arrays using unmanned aerial vehicles and deep learning

Published in Submitting, 1997

Submitting for mapping the solar panels using UAV imagery. We flew a DJI mini 2 taking imagery of solar panels similar to sizes of the solar home systems and estimate the cost of collecting drone imagery and compare them with the satellite imagery from private companies

Recommended citation: Ren, S., Hu, W., ..., Fetter, R., Bradbury, K., Malof, J. (2021). Mapping solar photovoltaic arrays using unmanned aerial vehicles and deep learning