Learning in an Uncertain World: Representing Ambiguity Through Multiple Hypotheses
Christian Rupprecht,Christian Rupprecht,Iro Laina,Robert S. DiPietro,Maximilian Baust +4 more
- pp 3611-3620
Reads0
Chats0
TLDR
This work proposes a frame-work for reformulating existing single-prediction models as multiple hypothesis prediction (MHP) models and an associated meta loss and optimization procedure to train them, and finds that MHP models outperform their single-hypothesis counterparts in all cases and expose valuable insights into the variability of predictions.Abstract:
Many prediction tasks contain uncertainty. In some cases, uncertainty is inherent in the task itself. In future prediction, for example, many distinct outcomes are equally valid. In other cases, uncertainty arises from the way data is labeled. For example, in object detection, many objects of interest often go unlabeled, and in human pose estimation, occluded joints are often labeled with ambiguous values. In this work we focus on a principled approach for handling such scenarios. In particular, we propose a frame-work for reformulating existing single-prediction models as multiple hypothesis prediction (MHP) models and an associated meta loss and optimization procedure to train them. To demonstrate our approach, we consider four diverse applications: human pose estimation, future prediction, image classification and segmentation. We find that MHP models outperform their single-hypothesis counterparts in all cases, and that MHP models simultaneously expose valuable insights into the variability of predictions.read more
Citations
More filters
Proceedings ArticleDOI
Multimodal Trajectory Predictions for Autonomous Driving using Deep Convolutional Networks
Henggang Cui,Vladan Radosavljevic,Fang-Chieh Chou,Tsung-Han Lin,Thi Nguyen,Tzu-Kuo Huang,Jeff Schneider,Nemanja Djuric +7 more
TL;DR: This work presents a method to predict multiple possible trajectories of actors while also estimating their probabilities, and successfully tested on SDVs in closed-course tests.
Posted Content
A Probabilistic U-Net for Segmentation of Ambiguous Images
Simon A. A. Kohl,Bernardino Romera-Paredes,Clemens Meyer,Jeffrey De Fauw,Joseph R. Ledsam,Klaus H. Maier-Hein,S. M. Ali Eslami,Danilo Jimenez Rezende,Olaf Ronneberger +8 more
TL;DR: A generative segmentation model based on a combination of a U-Net with a conditional variational autoencoder that is capable of efficiently producing an unlimited number of plausible hypotheses and reproduces the possible segmentation variants as well as the frequencies with which they occur significantly better than published approaches.
Proceedings ArticleDOI
UC-Net: Uncertainty Inspired RGB-D Saliency Detection via Conditional Variational Autoencoders
TL;DR: Zhang et al. as mentioned in this paper proposed a probabilistic RGB-D saliency detection network via conditional variational autoencoders to model human annotation uncertainty and generate multiple saliency maps for each input image by sampling in the latent space.
Posted Content
UC-Net: Uncertainty Inspired RGB-D Saliency Detection via Conditional Variational Autoencoders
TL;DR: Inspired by the saliency data labeling process, a probabilistic RGB-D saliency detection network via conditional variational autoencoders to model human annotation uncertainty and generate multiple saliency maps for each input image by sampling in the latent space is proposed.
Proceedings ArticleDOI
ContactDB: Analyzing and Predicting Grasp Contact via Thermal Imaging
TL;DR: This work presents ContactDB, a novel dataset of contact maps for household objects that captures the rich hand-object contact that occurs during grasping, enabled by use of a thermal camera.
References
More filters
Book ChapterDOI
An Uncertain Future: Forecasting from Static Images Using Variational Autoencoders
TL;DR: In this article, a conditional variational autoencoder is proposed to predict the dense trajectory of pixels in a scene, where it will travel, and how it will deform over the course of one second.
Journal ArticleDOI
Deep Label Distribution Learning With Label Ambiguity
TL;DR: The proposed deep label distribution learning (DLDL) method effectively utilizes the label ambiguity in both feature learning and classifier learning, which help prevent the network from overfitting even when the training set is small.
Posted Content
Deep Convolutional Ranking for Multilabel Image Annotation
TL;DR: In this paper, a significant performance gain could be obtained by combining convolutional architectures with approximate top-k$ ranking objectives, as they naturally fit the multilabel tagging problem.
Posted Content
Learning Physical Intuition of Block Towers by Example
Adam Lerer,Sam Gross,Rob Fergus +2 more
TL;DR: This paper creates small towers of wooden blocks whose stability is randomized and render them collapsing (or remaining upright) to train large convolutional network models which can accurately predict the outcome, as well as estimating the block trajectories.
Proceedings ArticleDOI
Robust Optimization for Deep Regression
TL;DR: In this article, the authors proposed a regression model with ConvNets that achieves robustness to such outliers by minimizing Tukey's biweight function, an M-estimator robust to outliers, as the loss function for the ConvNet.