Search Results for author: Niko Sünderhauf

Found 38 papers, 14 papers with code

Addressing the Challenges of Open-World Object Detection

no code implementations27 Mar 2023 David Pershouse, Feras Dayoub, Dimity Miller, Niko Sünderhauf

We address the challenging problem of open world object detection (OWOD), where object detectors must identify objects from known classes while also identifying and continually learning to detect novel objects.

object-detection Open World Object Detection

ParticleNeRF: A Particle-Based Encoding for Online Neural Radiance Fields

no code implementations8 Nov 2022 Jad Abou-Chakra, Feras Dayoub, Niko Sünderhauf

ParticleNeRF is the first online dynamic NeRF and achieves fast adaptability with better visual fidelity than brute-force online InstantNGP and other baseline approaches on dynamic scenes with online constraints.

Residual Skill Policies: Learning an Adaptable Skill-based Action Space for Reinforcement Learning for Robotics

1 code implementation4 Nov 2022 Krishan Rana, Ming Xu, Brendan Tidd, Michael Milford, Niko Sünderhauf

Furthermore, the downstream RL agent is limited to learning structurally similar tasks to those used to construct the skill space.

Reinforcement Learning (RL)

Density-aware NeRF Ensembles: Quantifying Predictive Uncertainty in Neural Radiance Fields

no code implementations19 Sep 2022 Niko Sünderhauf, Jad Abou-Chakra, Dimity Miller

We show that ensembling effectively quantifies model uncertainty in Neural Radiance Fields (NeRFs) if a density-aware epistemic uncertainty term is considered.

Hyperdimensional Feature Fusion for Out-Of-Distribution Detection

1 code implementation10 Dec 2021 Samuel Wilson, Tobias Fischer, Niko Sünderhauf, Feras Dayoub

We introduce powerful ideas from Hyperdimensional Computing into the challenging field of Out-of-Distribution (OOD) detection.

Out-of-Distribution Detection Out of Distribution (OOD) Detection

Zero-Shot Uncertainty-Aware Deployment of Simulation Trained Policies on Real-World Robots

no code implementations10 Dec 2021 Krishan Rana, Vibhavari Dasagi, Jesse Haviland, Ben Talbot, Michael Milford, Niko Sünderhauf

While deep reinforcement learning (RL) agents have demonstrated incredible potential in attaining dexterous behaviours for robotics, they tend to make errors when deployed in the real world due to mismatches between the training and execution environments.

Continuous Control Reinforcement Learning (RL)

FSNet: A Failure Detection Framework for Semantic Segmentation

no code implementations19 Aug 2021 Quazi Marufur Rahman, Niko Sünderhauf, Peter Corke, Feras Dayoub

Semantic segmentation is an important task that helps autonomous vehicles understand their surroundings and navigate safely.

Autonomous Vehicles Navigate +1

Bayesian Controller Fusion: Leveraging Control Priors in Deep Reinforcement Learning for Robotics

no code implementations21 Jul 2021 Krishan Rana, Vibhavari Dasagi, Jesse Haviland, Ben Talbot, Michael Milford, Niko Sünderhauf

More importantly, given the risk-aversity of the control prior, BCF ensures safe exploration and deployment, where the control prior naturally dominates the action distribution in states unknown to the policy.

reinforcement-learning Reinforcement Learning (RL) +1

Probabilistic Appearance-Invariant Topometric Localization with New Place Awareness

1 code implementation16 Jul 2021 Ming Xu, Tobias Fischer, Niko Sünderhauf, Michael Milford

Probabilistic state-estimation approaches offer a principled foundation for designing localization systems, because they naturally integrate sequences of imperfect motion and exteroceptive sensor data.

Loop Closure Detection Visual Place Recognition

Probabilistic Visual Place Recognition for Hierarchical Localization

1 code implementation7 May 2021 Ming Xu, Niko Sünderhauf, Michael Milford

In this letter, we propose two methods which adapt image retrieval techniques used for visual place recognition to the Bayesian state estimation formulation for localization.

Image Retrieval Retrieval +2

Uncertainty for Identifying Open-Set Errors in Visual Object Detection

1 code implementation3 Apr 2021 Dimity Miller, Niko Sünderhauf, Michael Milford, Feras Dayoub

We also introduce a methodology for converting existing object detection datasets into specific open-set datasets to evaluate open-set performance in object detection.

object-detection Object Detection

Semantics for Robotic Mapping, Perception and Interaction: A Survey

no code implementations2 Jan 2021 Sourav Garg, Niko Sünderhauf, Feras Dayoub, Douglas Morrison, Akansel Cosgun, Gustavo Carneiro, Qi Wu, Tat-Jun Chin, Ian Reid, Stephen Gould, Peter Corke, Michael Milford

In robotics and related research fields, the study of understanding is often referred to as semantics, which dictates what does the world "mean" to a robot, and is strongly tied to the question of how to represent that meaning.

Autonomous Driving Navigate

SWA Object Detection

2 code implementations23 Dec 2020 Haoyang Zhang, Ying Wang, Feras Dayoub, Niko Sünderhauf

In this technique report, we systematically investigate the effects of applying SWA to object detection as well as instance segmentation.

Instance Segmentation object-detection +2

Online Monitoring of Object Detection Performance During Deployment

no code implementations16 Nov 2020 Quazi Marufur Rahman, Niko Sünderhauf, Feras Dayoub

During deployment, an object detector is expected to operate at a similar performance level reported on its testing dataset.

Autonomous Driving object-detection +1

Per-frame mAP Prediction for Continuous Performance Monitoring of Object Detection During Deployment

1 code implementation18 Sep 2020 Quazi Marufur Rahman, Niko Sünderhauf, Feras Dayoub

Performance monitoring of object detection is crucial for safety-critical applications such as autonomous vehicles that operate under varying and complex environmental conditions.

Autonomous Vehicles object-detection +1

VarifocalNet: An IoU-aware Dense Object Detector

2 code implementations CVPR 2021 Haoyang Zhang, Ying Wang, Feras Dayoub, Niko Sünderhauf

In this paper, we propose to learn an Iou-aware Classification Score (IACS) as a joint representation of object presence confidence and localization accuracy.

General Classification

BenchBot: Evaluating Robotics Research in Photorealistic 3D Simulation and on Real Robots

no code implementations3 Aug 2020 Ben Talbot, David Hall, Haoyang Zhang, Suman Raj Bista, Rohan Smith, Feras Dayoub, Niko Sünderhauf

We introduce BenchBot, a novel software suite for benchmarking the performance of robotics research across both photorealistic 3D simulations and real robot platforms.


Class Anchor Clustering: a Loss for Distance-based Open Set Recognition

1 code implementation6 Apr 2020 Dimity Miller, Niko Sünderhauf, Michael Milford, Feras Dayoub

We also show that our anchored class centres achieve higher open set performance than learnt class centres, particularly on object-based datasets and large numbers of training classes.

Open Set Learning

Multiplicative Controller Fusion: Leveraging Algorithmic Priors for Sample-efficient Reinforcement Learning and Safe Sim-To-Real Transfer

1 code implementation11 Mar 2020 Krishan Rana, Vibhavari Dasagi, Ben Talbot, Michael Milford, Niko Sünderhauf

We present a novel approach to model-free reinforcement learning that can leverage existing sub-optimal solutions as an algorithmic prior during training and deployment.

Robot Navigation

Residual Reactive Navigation: Combining Classical and Learned Navigation Strategies For Deployment in Unknown Environments

no code implementations24 Sep 2019 Krishan Rana, Ben Talbot, Vibhavari Dasagi, Michael Milford, Niko Sünderhauf

In this work we focus on improving the efficiency and generalisation of learned navigation strategies when transferred from its training environment to previously unseen ones.

Where are the Keys? -- Learning Object-Centric Navigation Policies on Semantic Maps with Graph Convolutional Networks

1 code implementation16 Sep 2019 Niko Sünderhauf

Emerging object-based SLAM algorithms can build a graph representation of an environment comprising nodes for robot poses and object landmarks.

Semantic Similarity Semantic Textual Similarity

The Probabilistic Object Detection Challenge

no code implementations19 Mar 2019 John Skinner, David Hall, Haoyang Zhang, Feras Dayoub, Niko Sünderhauf

We introduce a new challenge for computer and robotic vision, the first ACRV Robotic Vision Challenge, Probabilistic Object Detection.

object-detection Object Detection

Probabilistic Object Detection: Definition and Evaluation

1 code implementation27 Nov 2018 David Hall, Feras Dayoub, John Skinner, Haoyang Zhang, Dimity Miller, Peter Corke, Gustavo Carneiro, Anelia Angelova, Niko Sünderhauf

We introduce Probabilistic Object Detection, the task of detecting objects in images and accurately quantifying the spatial and semantic uncertainties of the detections.

object-detection Object Detection

Sim-to-Real Transfer of Robot Learning with Variable Length Inputs

no code implementations20 Sep 2018 Vibhavari Dasagi, Robert Lee, Serena Mou, Jake Bruce, Niko Sünderhauf, Jürgen Leitner

Current end-to-end deep Reinforcement Learning (RL) approaches require jointly learning perception, decision-making and low-level control from very sparse reward signals and high-dimensional inputs, with little capability of incorporating prior knowledge.

Decision Making object-detection +4

The Limits and Potentials of Deep Learning for Robotics

no code implementations18 Apr 2018 Niko Sünderhauf, Oliver Brock, Walter Scheirer, Raia Hadsell, Dieter Fox, Jürgen Leitner, Ben Upcroft, Pieter Abbeel, Wolfram Burgard, Michael Milford, Peter Corke

In this paper we discuss a number of robotics-specific learning, reasoning, and embodiment challenges for deep learning.


QuadricSLAM: Dual Quadrics from Object Detections as Landmarks in Object-oriented SLAM

no code implementations10 Apr 2018 Lachlan Nicholson, Michael Milford, Niko Sünderhauf

In this paper, we use 2D object detections from multiple views to simultaneously estimate a 3D quadric surface for each object and localize the camera position.


Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environments

7 code implementations CVPR 2018 Peter Anderson, Qi Wu, Damien Teney, Jake Bruce, Mark Johnson, Niko Sünderhauf, Ian Reid, Stephen Gould, Anton Van Den Hengel

This is significant because a robot interpreting a natural-language navigation instruction on the basis of what it sees is carrying out a vision and language process that is similar to Visual Question Answering.

Translation Vision and Language Navigation +3

Dropout Sampling for Robust Object Detection in Open-Set Conditions

no code implementations18 Oct 2017 Dimity Miller, Lachlan Nicholson, Feras Dayoub, Niko Sünderhauf

Dropout Variational Inference, or Dropout Sampling, has been recently proposed as an approximation technique for Bayesian Deep Learning and evaluated for image classification and regression tasks.

Image Classification object-detection +2

SceneCut: Joint Geometric and Object Segmentation for Indoor Scenes

no code implementations21 Sep 2017 Trung Pham, Thanh-Toan Do, Niko Sünderhauf, Ian Reid

This paper presents SceneCut, a novel approach to jointly discover previously unseen objects and non-object surfaces using a single RGB-D image.

Semantic Segmentation

Multi-Modal Trip Hazard Affordance Detection On Construction Sites

no code implementations21 Jun 2017 Sean McMahon, Niko Sünderhauf, Ben Upcroft, Michael Milford

Trip hazards are a significant contributor to accidents on construction and manufacturing sites, where over a third of Australian workplace injuries occur [1].

Affordance Detection

Episode-Based Active Learning with Bayesian Neural Networks

no code implementations21 Mar 2017 Feras Dayoub, Niko Sünderhauf, Peter Corke

We investigate different strategies for active learning with Bayesian deep neural networks.

Active Learning

Meaningful Maps With Object-Oriented Semantic Mapping

no code implementations26 Sep 2016 Niko Sünderhauf, Trung T. Pham, Yasir Latif, Michael Milford, Ian Reid

For intelligent robots to interact in meaningful ways with their environment, they must understand both the geometric and semantic properties of the scene surrounding them.


On the Performance of ConvNet Features for Place Recognition

1 code implementation17 Jan 2015 Niko Sünderhauf, Feras Dayoub, Sareh Shirazi, Ben Upcroft, Michael Milford

Computer vision datasets are very different in character to robotic camera data, real-time performance is essential, and performance priorities can be different.

Visual Navigation

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