Search Results for author: David Forsyth

Found 39 papers, 12 papers with code

How to Steer Your Adversary: Targeted and Efficient Model Stealing Defenses with Gradient Redirection

no code implementations28 Jun 2022 Mantas Mazeika, Bo Li, David Forsyth

To meet these challenges, we present a new approach to model stealing defenses called gradient redirection.

Long Scale Error Control in Low Light Image and Video Enhancement Using Equivariance

no code implementations2 Jun 2022 Sara Aghajanzadeh, David Forsyth

In this paper, we show that respecting equivariance -- the color of a restored pixel should be the same, however the image is cropped -- produces real improvements over the state of the art for restoration.

Quantization Video Enhancement +1

StyleGAN of All Trades: Image Manipulation with Only Pretrained StyleGAN

1 code implementation2 Nov 2021 Min Jin Chong, Hsin-Ying Lee, David Forsyth

Recently, StyleGAN has enabled various image manipulation and editing tasks thanks to the high-quality generation and the disentangled latent space.

Image Manipulation Image-to-Image Translation +1

Controlled GAN-Based Creature Synthesis via a Challenging Game Art Dataset -- Addressing the Noise-Latent Trade-Off

no code implementations19 Aug 2021 Vaibhav Vavilala, David Forsyth

While noise inputs to StyleGAN2 are essential for good synthesis, we find that coarse-scale noise interferes with latent variables on this dataset because both control long-scale image effects.

LSD-StructureNet: Modeling Levels of Structural Detail in 3D Part Hierarchies

no code implementations ICCV 2021 Dominic Roberts, Ara Danielyan, Hang Chu, Mani Golparvar-Fard, David Forsyth

Generative models for 3D shapes represented by hierarchies of parts can generate realistic and diverse sets of outputs.

Retrieve in Style: Unsupervised Facial Feature Transfer and Retrieval

1 code implementation ICCV 2021 Min Jin Chong, Wen-Sheng Chu, Abhishek Kumar, David Forsyth

We present Retrieve in Style (RIS), an unsupervised framework for facial feature transfer and retrieval on real images.

Disentanglement

GANs N' Roses: Stable, Controllable, Diverse Image to Image Translation (works for videos too!)

2 code implementations11 Jun 2021 Min Jin Chong, David Forsyth

This adversarial loss guarantees the map is diverse -- a very wide range of anime can be produced from a single content code.

Image-to-Image Translation Translation

Toward Accurate and Realistic Virtual Try-on Through Shape Matching and Multiple Warps

no code implementations22 Mar 2020 Kedan Li, Min Jin Chong, Jingen Liu, David Forsyth

However, obtaining a realistic image is challenging because the kinematics of garments is complex and because outline, texture, and shading cues in the image reveal errors to human viewers.

Image Generation Virtual Try-on

Exposing and Correcting the Gender Bias in Image Captioning Datasets and Models

no code implementations2 Dec 2019 Shruti Bhargava, David Forsyth

Interestingly, the predictions by this model on images with no humans, are also visibly different from the one trained on gendered captions.

Image Captioning

Effectively Unbiased FID and Inception Score and where to find them

1 code implementation CVPR 2020 Min Jin Chong, David Forsyth

In turn, this effectively bias-free estimate requires good estimates of scores with a finite number of samples.

Improving Style Transfer with Calibrated Metrics

1 code implementation21 Oct 2019 Mao-Chuang Yeh, Shuai Tang, Anand Bhattad, Chuhang Zou, David Forsyth

Style transfer methods produce a transferred image which is a rendering of a content image in the manner of a style image.

Style Transfer

Counterfactual Depth from a Single RGB Image

1 code implementation3 Sep 2019 Theerasit Issaranon, Chuhang Zou, David Forsyth

We describe a method that predicts, from a single RGB image, a depth map that describes the scene when a masked object is removed - we call this "counterfactual depth" that models hidden scene geometry together with the observations.

Coherent and Controllable Outfit Generation

no code implementations17 Jun 2019 Kedan Li, Chen Liu, David Forsyth

A user study suggests that people understand the match between the queries and the outfits produced by our method.

General Classification

Max-Sliced Wasserstein Distance and its use for GANs

no code implementations CVPR 2019 Ishan Deshpande, Yuan-Ting Hu, Ruoyu Sun, Ayis Pyrros, Nasir Siddiqui, Sanmi Koyejo, Zhizhen Zhao, David Forsyth, Alexander Schwing

Generative adversarial nets (GANs) and variational auto-encoders have significantly improved our distribution modeling capabilities, showing promise for dataset augmentation, image-to-image translation and feature learning.

Image-to-Image Translation Translation

An Approximate Shading Model with Detail Decomposition for Object Relighting

no code implementations20 Apr 2018 Zicheng Liao, Kevin Karsch, Hongyi Zhang, David Forsyth

We present an object relighting system that allows an artist to select an object from an image and insert it into a target scene.

Learning Type-Aware Embeddings for Fashion Compatibility

1 code implementation ECCV 2018 Mariya I. Vasileva, Bryan A. Plummer, Krishna Dusad, Shreya Rajpal, Ranjitha Kumar, David Forsyth

Outfits in online fashion data are composed of items of many different types (e. g. top, bottom, shoes) that share some stylistic relationship with one another.

Detecting Anomalous Faces with 'No Peeking' Autoencoders

no code implementations15 Feb 2018 Anand Bhattad, Jason Rock, David Forsyth

We describe a method for detecting an anomalous face image that meets these requirements.

Standard detectors aren't (currently) fooled by physical adversarial stop signs

no code implementations9 Oct 2017 Jiajun Lu, Hussein Sibai, Evan Fabry, David Forsyth

Finally, an adversarial pattern on a physical object that could fool a detector would have to be adversarial in the face of a wide family of parametric distortions (scale; view angle; box shift inside the detector; illumination; and so on).

Adversarial Attack

NO Need to Worry about Adversarial Examples in Object Detection in Autonomous Vehicles

no code implementations12 Jul 2017 Jiajun Lu, Hussein Sibai, Evan Fabry, David Forsyth

Instead, a trained neural network classifies most of the pictures taken from different distances and angles of a perturbed image correctly.

Autonomous Vehicles object-detection +1

SafetyNet: Detecting and Rejecting Adversarial Examples Robustly

no code implementations ICCV 2017 Jiajun Lu, Theerasit Issaranon, David Forsyth

SceneProof applies to images captured with depth maps (RGBD images) and checks if a pair of image and depth map is consistent.

Learning Diverse Image Colorization

1 code implementation CVPR 2017 Aditya Deshpande, Jiajun Lu, Mao-Chuang Yeh, Min Jin Chong, David Forsyth

Finally, we build a conditional model for the multi-modal distribution between grey-level image and the color field embeddings.

Colorization

Authoring image decompositions with generative models

no code implementations5 Dec 2016 Jason Rock, Theerasit Issaranon, Aditya Deshpande, David Forsyth

Instead, we can learn to decompose an image into layers that are "like this" by authoring generative models for each layer using proxy examples that capture the Platonic ideal (Mondrian images for albedo; rendered 3D primitives for shading; material swatches for shading detail).

CDVAE: Co-embedding Deep Variational Auto Encoder for Conditional Variational Generation

no code implementations1 Dec 2016 Jiajun Lu, Aditya Deshpande, David Forsyth

Such a model is difficult to train, because we do not usually have training data containing many different shadings for the same image.

Image Relighting

Learning to Localize Little Landmarks

no code implementations CVPR 2016 Saurabh Singh, Derek Hoiem, David Forsyth

We describe a method to find such landmarks by finding a sequence of latent landmarks, each with a prediction model.

Swapout: Learning an ensemble of deep architectures

no code implementations NeurIPS 2016 Saurabh Singh, Derek Hoiem, David Forsyth

When viewed as a regularization method swapout not only inhibits co-adaptation of units in a layer, similar to dropout, but also across network layers.

Learning Large-Scale Automatic Image Colorization

no code implementations ICCV 2015 Aditya Deshpande, Jason Rock, David Forsyth

The coefficients of the objective function are conditioned on image features, using a random forest.

Colorization

Sparse Depth Super Resolution

no code implementations CVPR 2015 Jiajun Lu, David Forsyth

We describe a method to produce detailed high resolution depth maps from aggressively subsampled depth measurements.

Semantic Segmentation Super-Resolution

Learning a Sequential Search for Landmarks

no code implementations CVPR 2015 Saurabh Singh, Derek Hoiem, David Forsyth

We propose a general method to find landmarks in images of objects using both appearance and spatial context.

An Approximate Shading Model for Object Relighting

no code implementations CVPR 2015 Zicheng Liao, Kevin Karsch, David Forsyth

With this object model, we build an object relighting system that allows an artist to select an object from an image and insert it into a 3D scene.

Fast Template Evaluation with Vector Quantization

no code implementations NeurIPS 2013 Mohammad Amin Sadeghi, David Forsyth

Our procedure allows speed and accuracy to be traded off in two ways: by choosing the number of Vector Quantization levels, and by choosing to rescore windows or not.

object-detection Object Detection +1

Cannot find the paper you are looking for? You can Submit a new open access paper.