Forrest Iandola

PhD in EECS, University of California, Berkeley, 2016.
B.S. in Computer Science, University of Illinois at Urbana-Champaign, 2012.
Illinois Math and Science Academy, 2008.

Forrest Iandola completed a PhD in EECS at UC Berkeley, where his research focused on deep neural networks. His best-known work includes deep learning infrastructure such as FireCaffe and deep models such as SqueezeNet and SqueezeDet. His advances in deep learning led to the founding of DeepScale, where he was CEO from 2015 to 2019. Forrest joined Tesla Autopilot in 2019, where he is a Sr Staff Machine Learning Scientist.

Forrest Iandola

Research and Industry Experience

Tesla, Palo Alto, CA
Sr Staff Machine Learning Scientist, October 2019 to present
  • Working in the Autopilot team.

DeepScale, Mountain View, CA
Co-founder and CEO, September 2015 to September 2019
  • Led a team of 30+ people to develop computationally-efficient computer vision and perception software for road vehicles.
  • We recruited some of the world's best engineers in deep learning and high-performance computing.
  • Raised 3 rounds of venture funding.

University of California, Berkeley, CA
Graduate Student Researcher, August 2012 to December 2016
  • Worked with Kurt Keutzer on resource-efficient deep learning, ranging from training on hundreds of GPUs, to deploying on embedded platforms.
  • Looked at the full stack, from neural architecture, to software, to datasets, to system architecture.
  • Developed the popular SqueezeNet deep neural network architecture.

Microsoft Research, Redmond, WA
Research Intern, June 2014 to September 2014
  • Worked on distributed (multi-server) training of deep neural networks.
  • Worked on one of the early papers in image captioning.

NVIDIA Corporation, Santa Clara, CA
Software Engineering Intern, June 2012 to August 2012
  • Added kernel priority scheduling to the CUDA programming language. My co-authors and I added the cudaStreamCreateWithPriority() function to CUDA.

Circa 2006 - 2011: internships at SLAC National Accelerator Laboratory, Lawrence Livermore National Laboratory, State Farm Research Center, and Fermi National Accelerator Laboratory

Publications

Albert Shaw, Daniel Hunter, Forrest Iandola, Sammy Sidhu
SqueezeNAS: Fast neural architecture search for faster semantic segmentation
ICCV Neural Architects Workshop, 2019 (Paper on arXiv) (Code on GitHub) (bibtex) (News article in EE Times)

Paden Tomasello, Sammy Sidhu, Anting Shen, Matthew W. Moskewicz, Nobie Redmon, Gayatri Joshi, Romi Phadte, Paras Jain, Forrest Iandola
DSCnet: Replicating Lidar Point Clouds with Deep Sensor Cloning
CVPR Workshop on Autonomous Driving, 2019 (Paper on arXiv) (bibtex)

Forrest Iandola, Kurt Keutzer
Small Neural Nets Are Beautiful: Enabling Embedded Systems with Small Deep-Neural-Network Architectures
Keynote at ESWEEK, 2017 (Paper on arXiv) (bibtex)

Bichen Wu, Forrest Iandola, Peter H. Jin, Kurt Keutzer
SqueezeDet: Unified, Small, Low Power Fully Convolutional Neural Networks for Real-Time Object Detection for Autonomous Driving
CVPR Embedded Vision Workshop, 2017 (Paper on arXiv) (Code on GitHub) (bibtex)

Forrest Iandola
Exploring the Design Space of Deep Convolutional Neural Networks at Large Scale
UC Berkeley PhD Dissertation, 2016. (Dissertation on arXiv) (Slides) (bibtex)

Forrest N. Iandola, Song Han, Matthew W. Moskewicz, Khalid Ashraf, William J. Dally, and Kurt Keutzer
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and < 0.5MB model size
arXiv technical report, 2016 (Paper on arXiv) (Code on GitHub) (bibtex)
***Offshoots of SqueezeNet include SqueezeNext, SqueezeSeg, and SqueezeDet

Steena Monteiro, Forrest Iandola, and Daniel Wong
STOMP: Statistical Techniques for Optimizing and Modeling Performance of blocked sparse matrix vector multiplication
International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD), 2016 (Paper) (bibtex)

Matthew Moskewicz, Forrest Iandola, and Kurt Keutzer
Boda-RTC: Productive Generation of Portable, Efficient Code for Convolutional Neural Networks on Mobile Computing Platforms
IEEE International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob), 2016 (Paper on arXiv) (Code on GitHub)

Khalid Ashraf, Bichen Wu, Forrest N. Iandola, Matthew W. Moskewicz, Kurt Keutzer
Shallow Networks for High-Accuracy Road Object-Detection
arXiv technical report, 2016 (Paper on arXiv)

Forrest N. Iandola, Khalid Ashraf, Matthew W. Moskewicz, and Kurt Keutzer
FireCaffe: near-linear acceleration of deep neural network training on compute clusters
Computer Vision and Pattern Recognition (CVPR), 2016. (Paper on arXiv)

Forrest N. Iandola, Anting Shen, Peter Gao, and Kurt Keutzer
DeepLogo: Hitting Logo Recognition with the Deep Neural Network Hammer
arXiv technical report, October 2015. (Paper on arXiv) (Reference in MIT Technology Review) (News article on The Stack)

Hao Fang, Saurabh Gupta, Forrest Iandola, Rupesh Srivastava, Li Deng, Piotr Dollar, Jianfeng Gao, Xiaodong He, Margaret Mitchell, John C. Platt, C. Lawrence Zitnick, Geoffrey Zweig
From Captions to Visual Concepts and Back
Computer Vision and Pattern Recognition (CVPR), 2015. (Paper on arXiv)

Ross Girshick, Forrest Iandola, Trevor Darrell, and Jitendra Malik
Deformable Part Models are Convolutional Neural Networks
Computer Vision and Pattern Recognition (CVPR), 2015. (Paper on arXiv) (Code on GitHub)

Forrest Iandola, Matthew Moskewicz, Kurt Keutzer.
libHOG: Energy-Efficient Histogram of Oriented Gradient Computation
International Conference on Intelligent Transportation Systems (ITSC), 2015. (Paper) (Code on GitHub)

Khalid Ashraf, Benjamin Elizalde, Forrest Iandola, Matthew Moskewicz, Julia Bernd, Gerald Friedland, Kurt Keutzer.
Audio-Based Multimedia Event Detection with DNNs and Sparse Sampling
International Conference on Multimedia Retrieval (ICMR), 2015. (Paper) (Poster) (Pre-release version of code)

Forrest Iandola, Matt Moskewicz, Sergey Karayev, Ross Girshick, Trevor Darrell, Kurt Keutzer
DenseNet: Implementing Efficient ConvNet Descriptor Pyramids
arXiv technical report, April 2014. (Paper on arXiv) (Code in Caffe featpyra_int_rc branch on GitHub)

Michael Anderson, Forrest Iandola, and Kurt Keutzer
Quantifying the Energy Efficiency of Object Recognition and Optical Flow
UC Berkeley technical report, March 2014. (Paper)

Forrest N. Iandola, David Sheffield, Michael Anderson, Phitchaya Mangpo Phothilimthana, and Kurt Keutzer
Communication-Minimizing 2D Convolution in GPU Registers
IEEE International Conference on Image Processing (ICIP), 2013. (Paper) (Slides) (Code on GitHub)

Ning Zhang, Ryan Farrell, Forrest Iandola, and Trevor Darrell
Deformable Part Descriptors for Fine-grained Recognition and Attribute Prediction
International Conference on Computer Vision (ICCV), 2013. (Paper) (Project Page)

Mehdi Maasoumy, Pierluigi Nuzzo, Forrest Iandola, Maryam Kamgarpour, Alberto Sangiovanni-Vincentelli and Claire Tomlin
Optimal Load Management System for Aircraft Electric Power Distribution
IEEE Conference on Decision and Control (CDC), 2013. (Paper) (Code on GitHub)

Vivek Kini, Forrest Iandola, and Timothy Murray
Techniques for Assigning Priorities to Streams of Work
US Patent #US20140344822.

Forrest N. Iandola, Jan Schuemann, Jungwook Shin, Bruce Faddegon, Harald Paganetti, and Joseph Perl
Representing Range Compensators in the TOPAS Monte Carlo System
European Workshop on Monte Carlo Treatment Planning, 2012. (Paper) (Slides)

Fatemeh Saremi, Praveen Jayachandran, Forrest Iandola, Yusuf Sarwar, and Tarek Abdelzaher
On Schedulability and Time Composability of Multisensor Data Aggregation Networks
International Conference on Information Fusion, 2012. (Paper) (Slides)

Mahsa Kamali, Forrest N. Iandola, Hui Fang, and John C. Hart
MethMorph: Simulating Facial Deformation due to Methamphetamine Usage
International Symposium on Visual Computing (ISVC), 2011. (Paper)

Mahsa Kamali, Ido Omer, Forrest Iandola, Eyal Ofek, and John C. Hart
Linear Clutter Removal from Urban Panoramas
International Symposium on Visual Computing (ISVC), 2011. (Paper)

Forrest Iandola, Fatemeh Saremi, Tarek Abdelzaher, Praveen Jayachandran, and Aylin Yener
Real-Time Capacity of Networked Data Fusion
International Conference on Information Fusion, 2011. (Paper) (Slides)

Forrest N. Iandola, Matthew O'Brien, and Richard Procassini
PyMercury: Interactive Python for the Mercury Monte Carlo Particle Transport Code
International Conference on Mathematics and Computational Methods Applied to Nuclear Science and Engineering (M&C), 2011. (Paper) (Poster)

Forrest Iandola and Michael Syphers
Electron Beam Focusing for the International Linear Collider
American Association for the Advancement of Science Annual Meeting, 2008. (Paper)

Talks, Posters, and Projects

Albert Shaw, Daniel Hunter, Sammy Sidhu, Forrest Iandola
Squeezing down the computing requirements of deep neural networks
IEEE Santa Clara Valley Section, 2019 (Slides)

Albert Shaw, Daniel Hunter, Sammy Sidhu, Forrest Iandola
SqueezeNAS: Fast neural architecture search for faster semantic segmentation
CVPR Workshop on Autonomous Driving, 2019 (Invited Talk) (Slides)

Forrest Iandola, Kurt Keutzer, Joseph E. Gonzalez
Survey of Enabling Technologies for Automated Driving
Short Course at AutoSens, 2019 (Slides)

Forrest Iandola, Kurt Keutzer, Joseph E. Gonzalez
Developing Enabling Technologies for Automated Driving
Short Course at Electronic Imaging, 2019 (Slides) (Video)

Forrest Iandola
Tips and Tricks for Developing Smaller Neural Nets
CVPR Workshop on Efficient Deep Learning for Computer Vision, 2018 (Invited Talk) (Slides) (Video)

Forrest Iandola
Scaling the Training of Deep Neural Networks
Guest Lecture in CS267 at UC Berkeley, 2018 (Video)

Forrest Iandola and Kurt Keutzer
Small Deep-Neural-Networks: Their Advantages and Their Design
ICML TinyML Workshop, 2017 (Invited Talk) (Slides) (Video) (Blog Post)

Forrest Iandola
Distributed deep neural network training: A measurement study
UC Berkeley CS268 class project, May 2016 (Paper) (Slides)

Michael Anderson, Khalid Ashraf, Gerald Friedland, Forrest Iandola, Peter Jin, Matt Moskewicz, Zach Rowisnki, Kurt Keutzer
Application-Driven Research in the ASPIRE Lab
ASPIRE Retreat, May 2014 (Slides)

Forrest Iandola, Matt Moskewicz, Sergey Karayev, Ross Girshick, Kurt Keutzer, and Trevor Darrell
DenseNet: Efficient Computation of Deep Neural Networks for Object Detection
Presented versions of this at BVLC and ASPIRE retreats, March-May 2014. (Poster)

Matt Moskewicz, Forrest Iandola, and Kurt Keutzer
Boda: Towards a Framework for Evaluating Accuracy/Efficiency Tradeoffs in Object Detection
ASPIRE Retreat, Jan 2014 (Poster)

Forrest Iandola, Ning Zhang, Ross Girshick, Trevor Darrell, and Kurt Keutzer
Scaling Up Deformable Parts Models for Object Detection
ASPIRE Retreat, May 2013 (Poster)

Forrest Iandola, David Sheffield, Michael Anderson, Mangpo Phothilimthana, and Kurt Keutzer
Minimizing Memory Communication for 2D Image Convolution in GPU Registers
GPU Technology Conference, March 2013 (Poster)
   

Contact

Email: forrest.dnn@gmail.com

My Blog

My Old Blog

Forrest Iandola on LinkedIn Forrest Iandola on Twitter Forrest Iandola on Facebook