Scott Reed


I completed my PhD at the University of Michigan.

My adviser was Honglak Lee.

Currently I am a research scientist at Google DeepMind.

Contact: reedscot at the usual umich dot edu address.
Linkedin: linkedin.com/in/reedscott
GitHub: github.com/reedscot


Interests

My current research topics are related to representation learning and deep learning, scalable object detection, multimodal learning, learning to disentangle factors of variation in sensory data, visual analogy-making, and learning compositional programs. Broadly I am interested in neural architectures that can automatically learn feature representations with strong generalization properties in a data-efficient manner.


Publications

Generating Interpretable Images with Controllable Structure [Preprint PDF][BibTex]
Scott Reed, AƤron van den Oord, Nal Kalchbrenner, Victor Bapst, Matt Botvinick, Nando de Freitas

Learning What and Where to Draw [PDF][BibTex][Code and Data][Supplement]
Scott Reed, Zeynep Akata, Santosh Mohan, Samuel Tenka, Bernt Schiele, Honglak Lee
Oral Presentation
In Advances in Neural Information Processing Systems (NIPS), Barcelona, 2016.

Generative Adversarial Text-to-Image Synthesis [PDF][Supplement][BibTex][Code]
Scott Reed, Zeynep Akata, Xinchen Yan, Lajanugen Logeswaran, Bernt Schiele, Honglak Lee.
In Proceedings of the 33rd International Conference on Machine Learning (ICML), New York, USA, 2016.

Neural Programmer-Interpreters [PDF][Slides][BibTex][Project page]
Scott Reed and Nando de Freitas
This work was done during my internship at Google DeepMind.
Best Paper Award
In International Conference on Learning Representations (ICLR), San Juan, Puerto Rico, 2016.

SSD: Single Shot MultiBox Detector [PDF][BibTex]
Wei Liu, Dragomir Anguelov, Dumitru Erhan, Christian Szegedy, Scott Reed, Cheng-Yang Fu, Alexander C. Berg
In European Conference on Computer Vision (ECCV), Amsterdam, 2016.

Learning Deep Representations of Fine-Grained Visual Descriptions [PDF][BibTex][Code]
Scott Reed, Zeynep Akata, Bernt Schiele, Honglak Lee
Spotlight Presentation
In IEEE Computer Vision and Pattern Recognition (CVPR), Las Vegas, USA, 2016.

Deep Visual Analogy-Making [PDF][BibTex][Slides][Code][Data]
Scott Reed, Yi Zhang, Yuting Zhang, Honglak Lee
Oral Presentation
In Advances in Neural Information Processing Systems (NIPS), Montreal, 2015.

Weakly-supervised Disentangling with Recurrent Transformations for 3D View Synthesis [PDF][BibTex]
Jimei Yang, Scott Reed, Ming-Hsuan Yang, Honglak Lee
In Advances in Neural Information Processing Systems (NIPS), Montreal, 2015.

Scalable, High-Quality Object Detection [PDF][BibTex]
Christian Szegedy, Scott Reed, Dumitru Erhan, Dragomir Anguelov

Evaluation of Output Embeddings for Fine-Grained Image Classification [PDF][BibTex]
Zeynep Akata, Scott Reed, Daniel Walter, Honglak Lee, Bernt Schiele
In IEEE Computer Vision and Pattern Recognition (CVPR), Boston, USA, 2015.

Going deeper with convolutions [PDF][BibTex]
Christian Szegedy, Wei Liu, Yangqing Jia, Pierre Sermanet, Scott Reed, Dragomir Anguelov, Dumitru Erhan, Vincent Vanhoucke, Andrew Rabinovich
In IEEE Computer Vision and Pattern Recognition (CVPR), Boston, USA, 2015.
Related Google Research Blog: Building a deeper understanding of images

Learning to Disentangle Factors of Variation with Manifold Interaction [PDF][Code][BibTex]
Scott Reed, Kihyuk Sohn, Yuting Zhang, Honglak Lee.
In Proceedings of the 31st International Conference on Machine Learning (ICML), Beijing, China, 2014.


Awards and Honors

National Science Foundation Graduate Research Fellowship Program (NSF GRFP)
National Defense Science and Engineering Graduate Fellowship (NDSEG)
Telluride Association Room and Board Scholarship (2007-2011)
Chrysler Foundation Scholarship (2007-2011)
Phi Beta Kappa (2011)
NSF SUBMERGE REU scholarship (2009-2011)