• Corpus ID: 49314531

Bayesian Prediction of Future Street Scenes through Importance Sampling based Optimization

@article{Bhattacharyya2018BayesianPO,
  title={Bayesian Prediction of Future Street Scenes through Importance Sampling based Optimization},
  author={Apratim Bhattacharyya and Mario Fritz and Bernt Schiele},
  journal={ArXiv},
  year={2018},
  volume={abs/1806.06939}
}
For autonomous agents to successfully operate in the real world, anticipation of future events and states of their environment is a key competence. This problem can be formalized as a sequence prediction problem, where a number of observations are used to predict the sequence into the future. However, real-world scenarios demand a model of uncertainty of such predictions, as future states become increasingly uncertain and multi-modal -- in particular on long time horizons. This makes modelling… 

Figures and Tables from this paper

Multimodal Future Localization and Emergence Prediction for Objects in Egocentric View With a Reachability Prior

TLDR
Experiments show that the reachability prior combined with multi-hypotheses learning improves multimodal prediction of the future location of tracked objects and, for the first time, the emergence of new objects.

Game Plan: What AI can do for Football, and What Football can do for AI

TLDR
An overarching perspective is provided highlighting how the combination of these fields, in particular, forms a unique microcosm for AI research, while offering mutual benefits for professional teams, spectators, and broadcasters in the years to come.

References

SHOWING 1-10 OF 28 REFERENCES

Accurate and Diverse Sampling of Sequences Based on a "Best of Many" Sample Objective

TLDR
This work addresses challenges in a Gaussian Latent Variable model for sequence prediction with a "Best of Many" sample objective that leads to more accurate and more diverse predictions that better capture the true variations in real-world sequence data.

DESIRE: Distant Future Prediction in Dynamic Scenes with Interacting Agents

TLDR
The proposed Deep Stochastic IOC RNN Encoder-decoder framework, DESIRE, for the task of future predictions of multiple interacting agents in dynamic scenes significantly improves the prediction accuracy compared to other baseline methods.

Stochastic Variational Video Prediction

TLDR
This paper develops a stochastic variational video prediction (SV2P) method that predicts a different possible future for each sample of its latent variables, and is the first to provide effective Stochastic multi-frame prediction for real-world video.

Long-Term On-board Prediction of People in Traffic Scenes Under Uncertainty

TLDR
It is argued that it is necessary to predict at least 1 second and a new model is proposed that jointly predicts ego motion and people trajectories over such large time horizons and it is shown that both sequence modeling of trajectories as well as the novel method of long term odometry prediction are essential for best performance.

Predicting Scene Parsing and Motion Dynamics in the Future

TLDR
A novel model to simultaneously predict scene parsing and optical flow in unobserved future video frames is proposed and shows significantly better parsing and motion prediction results when compared to well-established baselines and individual prediction models on the large-scale Cityscapes dataset.

Predicting Deeper into the Future of Semantic Segmentation

TLDR
An autoregressive convolutional neural network that learns to iteratively generate multiple frames is developed and results show that directly predicting future segmentations is substantially better than predicting and then segmenting future RGB frames.

Learning Structured Output Representation using Deep Conditional Generative Models

TLDR
A deep conditional generative model for structured output prediction using Gaussian latent variables is developed, trained efficiently in the framework of stochastic gradient variational Bayes, and allows for fast prediction using Stochastic feed-forward inference.

Visual Dynamics: Probabilistic Future Frame Synthesis via Cross Convolutional Networks

TLDR
A novel approach that models future frames in a probabilistic manner is proposed, namely a Cross Convolutional Network to aid in synthesizing future frames; this network structure encodes image and motion information as feature maps and convolutional kernels, respectively.

Predicting Future Instance Segmentations by Forecasting Convolutional Features

TLDR
A predictive model is developed in the space of fixed-sized convolutional features of the Mask R-CNN instance segmentation model that significantly improves over baselines based on optical flow.

Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning

TLDR
A new theoretical framework is developed casting dropout training in deep neural networks (NNs) as approximate Bayesian inference in deep Gaussian processes, which mitigates the problem of representing uncertainty in deep learning without sacrificing either computational complexity or test accuracy.