Search Results for author: Ingmar Posner

Found 65 papers, 20 papers with code

Gaitor: Learning a Unified Representation Across Gaits for Real-World Quadruped Locomotion

no code implementations29 May 2024 Alexander L. Mitchell, Wolfgang Merkt, Aristotelis Papatheodorou, Ioannis Havoutis, Ingmar Posner

The current state-of-the-art in quadruped locomotion is able to produce robust motion for terrain traversal but requires the segmentation of a desired robot trajectory into a discrete set of locomotion skills such as trot and crawl.

Compete and Compose: Learning Independent Mechanisms for Modular World Models

no code implementations23 Apr 2024 Anson Lei, Frederik Nolte, Bernhard Schölkopf, Ingmar Posner

COMET is trained on multiple environments with varying dynamics via a two-step process: competition and composition.

D-Cubed: Latent Diffusion Trajectory Optimisation for Dexterous Deformable Manipulation

no code implementations19 Mar 2024 Jun Yamada, Shaohong Zhong, Jack Collins, Ingmar Posner

In this work, we propose D-Cubed, a novel trajectory optimisation method using a latent diffusion model (LDM) trained from a task-agnostic play dataset to solve dexterous deformable object manipulation tasks.

Deformable Object Manipulation

World Models via Policy-Guided Trajectory Diffusion

1 code implementation13 Dec 2023 Marc Rigter, Jun Yamada, Ingmar Posner

Our results demonstrate that PolyGRAD outperforms state-of-the-art baselines in terms of trajectory prediction error for short trajectories, with the exception of autoregressive diffusion.

Continuous Control Denoising +2

TWIST: Teacher-Student World Model Distillation for Efficient Sim-to-Real Transfer

no code implementations7 Nov 2023 Jun Yamada, Marc Rigter, Jack Collins, Ingmar Posner

The teacher world model then supervises a student world model that takes the domain-randomised image observations as input.

Gromov-Hausdorff Distances for Comparing Product Manifolds of Model Spaces

no code implementations9 Sep 2023 Haitz Saez de Ocariz Borde, Alvaro Arroyo, Ismael Morales, Ingmar Posner, Xiaowen Dong

Recent studies propose enhancing machine learning models by aligning the geometric characteristics of the latent space with the underlying data structure.

AutoGraph: Predicting Lane Graphs from Traffic Observations

1 code implementation27 Jun 2023 Jannik Zürn, Ingmar Posner, Wolfram Burgard

To overcome this limitation, we propose to use the motion patterns of traffic participants as lane graph annotations.

Autonomous Driving

Reward-Free Curricula for Training Robust World Models

1 code implementation15 Jun 2023 Marc Rigter, Minqi Jiang, Ingmar Posner

We consider robustness in terms of minimax regret over all environment instantiations and show that the minimax regret can be connected to minimising the maximum error in the world model across environment instances.

You Only Look at One: Category-Level Object Representations for Pose Estimation From a Single Example

no code implementations22 May 2023 Walter Goodwin, Ioannis Havoutis, Ingmar Posner

In this work, we present a method for achieving category-level pose estimation by inspection of just a single object from a desired category.

6D Pose Estimation Continual Learning +1

Projections of Model Spaces for Latent Graph Inference

no code implementations21 Mar 2023 Haitz Sáez de Ocáriz Borde, Álvaro Arroyo, Ingmar Posner

Graph Neural Networks leverage the connectivity structure of graphs as an inductive bias.

Inductive Bias

Efficient Skill Acquisition for Complex Manipulation Tasks in Obstructed Environments

no code implementations6 Mar 2023 Jun Yamada, Jack Collins, Ingmar Posner

In this work, we propose a system for efficient skill acquisition that leverages an object-centric generative model (OCGM) for versatile goal identification to specify a goal for MP combined with RL to solve complex manipulation tasks in obstructed environments.

Motion Planning Object +1

Leveraging Scene Embeddings for Gradient-Based Motion Planning in Latent Space

no code implementations6 Mar 2023 Jun Yamada, Chia-Man Hung, Jack Collins, Ioannis Havoutis, Ingmar Posner

Motion planning framed as optimisation in structured latent spaces has recently emerged as competitive with traditional methods in terms of planning success while significantly outperforming them in terms of computational speed.

Motion Planning

DITTO: Offline Imitation Learning with World Models

no code implementations6 Feb 2023 Branton DeMoss, Paul Duckworth, Nick Hawes, Ingmar Posner

We propose DITTO, an offline imitation learning algorithm which uses world models and on-policy reinforcement learning to addresses the problem of covariate shift, without access to an oracle or any additional online interactions.

Imitation Learning reinforcement-learning +1

Reaching Through Latent Space: From Joint Statistics to Path Planning in Manipulation

no code implementations21 Oct 2022 Chia-Man Hung, Shaohong Zhong, Walter Goodwin, Oiwi Parker Jones, Martin Engelcke, Ioannis Havoutis, Ingmar Posner

We present a novel approach to path planning for robotic manipulators, in which paths are produced via iterative optimisation in the latent space of a generative model of robot poses.

Variational Causal Dynamics: Discovering Modular World Models from Interventions

no code implementations22 Jun 2022 Anson Lei, Bernhard Schölkopf, Ingmar Posner

In doing so, VCD significantly extends the capabilities of the current state-of-the-art in latent world models while also comparing favourably in terms of prediction accuracy.

Causal Discovery Variational Inference

ObPose: Leveraging Pose for Object-Centric Scene Inference and Generation in 3D

no code implementations7 Jun 2022 Yizhe Wu, Oiwi Parker Jones, Ingmar Posner

We present ObPose, an unsupervised object-centric inference and generation model which learns 3D-structured latent representations from RGB-D scenes.

Inductive Bias Object +2

VAE-Loco: Versatile Quadruped Locomotion by Learning a Disentangled Gait Representation

no code implementations2 May 2022 Alexander L. Mitchell, Wolfgang Merkt, Mathieu Geisert, Siddhant Gangapurwala, Martin Engelcke, Oiwi Parker Jones, Ioannis Havoutis, Ingmar Posner

We evaluate our approach on two versions of the real ANYmal quadruped robots and demonstrate that our method achieves a continuous blend of dynamic trot styles whilst being robust and reactive to external perturbations.


Zero-Shot Category-Level Object Pose Estimation

1 code implementation7 Apr 2022 Walter Goodwin, Sagar Vaze, Ioannis Havoutis, Ingmar Posner

Object pose estimation is an important component of most vision pipelines for embodied agents, as well as in 3D vision more generally.

Object Pose Estimation

Next Steps: Learning a Disentangled Gait Representation for Versatile Quadruped Locomotion

no code implementations9 Dec 2021 Alexander L. Mitchell, Wolfgang Merkt, Mathieu Geisert, Siddhant Gangapurwala, Martin Engelcke, Oiwi Parker Jones, Ioannis Havoutis, Ingmar Posner

This encourages disentanglement such that application of a drive signal to a single dimension of the latent state induces holistic plans synthesising a continuous variety of trot styles.


Semantically Grounded Object Matching for Robust Robotic Scene Rearrangement

1 code implementation15 Nov 2021 Walter Goodwin, Sagar Vaze, Ioannis Havoutis, Ingmar Posner

Object rearrangement has recently emerged as a key competency in robot manipulation, with practical solutions generally involving object detection, recognition, grasping and high-level planning.

Language Modelling Object +3

Universal Approximation of Functions on Sets

no code implementations5 Jul 2021 Edward Wagstaff, Fabian B. Fuchs, Martin Engelcke, Michael A. Osborne, Ingmar Posner

We provide a theoretical analysis of Deep Sets which shows that this universal approximation property is only guaranteed if the model's latent space is sufficiently high-dimensional.

E(n) Equivariant Normalizing Flows

1 code implementation NeurIPS 2021 Victor Garcia Satorras, Emiel Hoogeboom, Fabian B. Fuchs, Ingmar Posner, Max Welling

This paper introduces a generative model equivariant to Euclidean symmetries: E(n) Equivariant Normalizing Flows (E-NFs).

GENESIS-V2: Inferring Unordered Object Representations without Iterative Refinement

2 code implementations NeurIPS 2021 Martin Engelcke, Oiwi Parker Jones, Ingmar Posner

Moreover, object representations are often inferred using RNNs which do not scale well to large images or iterative refinement which avoids imposing an unnatural ordering on objects in an image but requires the a priori initialisation of a fixed number of object representations.

Clustering Image Generation +6

Introspective Visuomotor Control: Exploiting Uncertainty in Deep Visuomotor Control for Failure Recovery

no code implementations22 Mar 2021 Chia-Man Hung, Li Sun, Yizhe Wu, Ioannis Havoutis, Ingmar Posner

To recover from high uncertainty cases, the robot monitors its uncertainty along a trajectory and explores possible actions in the state-action space to bring itself to a more certain state.

Imitation Learning Robot Manipulation

Iterative SE(3)-Transformers

2 code implementations26 Feb 2021 Fabian B. Fuchs, Edward Wagstaff, Justas Dauparas, Ingmar Posner

Motivated by this application, we implement an iterative version of the SE(3)-Transformer, an SE(3)-equivariant attention-based model for graph data.

Protein Structure Prediction

There and Back Again: Learning to Simulate Radar Data for Real-World Applications

no code implementations29 Nov 2020 Rob Weston, Oiwi Parker Jones, Ingmar Posner

Here we propose to learn a radar sensor model capable of synthesising faithful radar observations based on simulated elevation maps.

Reconstruction Bottlenecks in Object-Centric Generative Models

1 code implementation13 Jul 2020 Martin Engelcke, Oiwi Parker Jones, Ingmar Posner

A range of methods with suitable inductive biases exist to learn interpretable object-centric representations of images without supervision.

Object Object Discovery

Goal-Conditioned End-to-End Visuomotor Control for Versatile Skill Primitives

1 code implementation19 Mar 2020 Oliver Groth, Chia-Man Hung, Andrea Vedaldi, Ingmar Posner

Visuomotor control (VMC) is an effective means of achieving basic manipulation tasks such as pushing or pick-and-place from raw images.

Imitation Learning Meta-Learning +1

Localising Faster: Efficient and precise lidar-based robot localisation in large-scale environments

no code implementations4 Mar 2020 Li Sun, Daniel Adolfsson, Martin Magnusson, Henrik Andreasson, Ingmar Posner, Tom Duckett

More importantly, the Gaussian method (i. e. deep probabilistic localisation) and non-Gaussian method (i. e. MCL) can be integrated naturally via importance sampling.

Under the Radar: Learning to Predict Robust Keypoints for Odometry Estimation and Metric Localisation in Radar

2 code implementations29 Jan 2020 Dan Barnes, Ingmar Posner

This paper presents a self-supervised framework for learning to detect robust keypoints for odometry estimation and metric localisation in radar.

Radar odometry

Improving End-to-End Object Tracking Using Relational Reasoning

no code implementations ICLR 2020 Fabian B. Fuchs, Adam R. Kosiorek, Li Sun, Oiwi Parker Jones, Ingmar Posner

Relational reasoning, the ability to model interactions and relations between objects, is valuable for robust multi-object tracking and pivotal for trajectory prediction.

Multi-Object Tracking Object +2

Attention-Privileged Reinforcement Learning

no code implementations19 Nov 2019 Sasha Salter, Dushyant Rao, Markus Wulfmeier, Raia Hadsell, Ingmar Posner

Image-based Reinforcement Learning is known to suffer from poor sample efficiency and generalisation to unseen visuals such as distractors (task-independent aspects of the observation space).

reinforcement-learning Reinforcement Learning (RL)

Imagine That! Leveraging Emergent Affordances for 3D Tool Synthesis

no code implementations30 Sep 2019 Yizhe Wu, Sudhanshu Kasewa, Oliver Groth, Sasha Salter, Li Sun, Oiwi Parker Jones, Ingmar Posner

In this paper we explore the richness of information captured by the latent space of a vision-based generative model.

Imagine That! Leveraging Emergent Affordances for Tool Synthesis in Reaching Tasks

no code implementations25 Sep 2019 Yizhe Wu, Sudhanshu Kasewa, Oliver Groth, Sasha Salter, Li Sun, Oiwi Parker Jones, Ingmar Posner

In this paper we investigate an artificial agent's ability to perform task-focused tool synthesis via imagination.


Attention Privileged Reinforcement Learning for Domain Transfer

no code implementations25 Sep 2019 Sasha Salter, Dushyant Rao, Markus Wulfmeier, Raia Hadsell, Ingmar Posner

Applying reinforcement learning (RL) to physical systems presents notable challenges, given requirements regarding sample efficiency, safety, and physical constraints compared to simulated environments.

reinforcement-learning Reinforcement Learning (RL)

Masking by Moving: Learning Distraction-Free Radar Odometry from Pose Information

no code implementations9 Sep 2019 Dan Barnes, Rob Weston, Ingmar Posner

This paper presents an end-to-end radar odometry system which delivers robust, real-time pose estimates based on a learned embedding space free of sensing artefacts and distractor objects.

Radar odometry

The Oxford Radar RobotCar Dataset: A Radar Extension to the Oxford RobotCar Dataset

3 code implementations3 Sep 2019 Dan Barnes, Matthew Gadd, Paul Murcutt, Paul Newman, Ingmar Posner

In this paper we present The Oxford Radar RobotCar Dataset, a new dataset for researching scene understanding using Millimetre-Wave FMCW scanning radar data.

Robotics Signal Processing

Guiding Physical Intuition with Neural Stethoscopes

no code implementations ICLR 2019 Fabian Fuchs, Oliver Groth, Adam Kosiorek, Alex Bewley, Markus Wulfmeier, Andrea Vedaldi, Ingmar Posner

Using an adversarial stethoscope, the network is successfully de-biased, leading to a performance increase from 66% to 88%.

Physical Intuition

On the Limitations of Representing Functions on Sets

no code implementations25 Jan 2019 Edward Wagstaff, Fabian B. Fuchs, Martin Engelcke, Ingmar Posner, Michael Osborne

Recent work on the representation of functions on sets has considered the use of summation in a latent space to enforce permutation invariance.

Gaussian Processes

Dropout Distillation for Efficiently Estimating Model Confidence

no code implementations27 Sep 2018 Corina Gurau, Alex Bewley, Ingmar Posner

We also propose better calibration within the state of the art Faster R-CNN object detection framework and show, using the COCO dataset, that DDN helps train better calibrated object detectors.

General Classification Image Classification +2

Scrutinizing and De-Biasing Intuitive Physics with Neural Stethoscopes

no code implementations14 Jun 2018 Fabian B. Fuchs, Oliver Groth, Adam R. Kosiorek, Alex Bewley, Markus Wulfmeier, Andrea Vedaldi, Ingmar Posner

Conversely, training on an easy dataset where visual cues are positively correlated with stability, the baseline model learns a bias leading to poor performance on a harder dataset.

Sequential Attend, Infer, Repeat: Generative Modelling of Moving Objects

1 code implementation NeurIPS 2018 Adam R. Kosiorek, Hyunjik Kim, Ingmar Posner, Yee Whye Teh

It can reliably discover and track objects throughout the sequence of frames, and can also generate future frames conditioning on the current frame, thereby simulating expected motion of objects.

TACO: Learning Task Decomposition via Temporal Alignment for Control

1 code implementation ICML 2018 Kyriacos Shiarlis, Markus Wulfmeier, Sasha Salter, Shimon Whiteson, Ingmar Posner

Many advanced Learning from Demonstration (LfD) methods consider the decomposition of complex, real-world tasks into simpler sub-tasks.

Incremental Adversarial Domain Adaptation for Continually Changing Environments

no code implementations20 Dec 2017 Markus Wulfmeier, Alex Bewley, Ingmar Posner

Continuous appearance shifts such as changes in weather and lighting conditions can impact the performance of deployed machine learning models.

Generative Adversarial Network Unsupervised Domain Adaptation

Driven to Distraction: Self-Supervised Distractor Learning for Robust Monocular Visual Odometry in Urban Environments

no code implementations17 Nov 2017 Dan Barnes, Will Maddern, Geoffrey Pascoe, Ingmar Posner

We present a self-supervised approach to ignoring "distractors" in camera images for the purposes of robustly estimating vehicle motion in cluttered urban environments.

Monocular Visual Odometry

What Makes a Place? Building Bespoke Place Dependent Object Detectors for Robotics

no code implementations7 Aug 2017 Jeffrey Hawke, Alex Bewley, Ingmar Posner

This paper is about enabling robots to improve their perceptual performance through repeated use in their operating environment, creating local expert detectors fitted to the places through which a robot moves.

Autonomous Driving

Mutual Alignment Transfer Learning

no code implementations25 Jul 2017 Markus Wulfmeier, Ingmar Posner, Pieter Abbeel

Training robots for operation in the real world is a complex, time consuming and potentially expensive task.

Transfer Learning

Hierarchical Attentive Recurrent Tracking

1 code implementation NeurIPS 2017 Adam R. Kosiorek, Alex Bewley, Ingmar Posner

Class-agnostic object tracking is particularly difficult in cluttered environments as target specific discriminative models cannot be learned a priori.

Activity Recognition Object +1

Addressing Appearance Change in Outdoor Robotics with Adversarial Domain Adaptation

no code implementations4 Mar 2017 Markus Wulfmeier, Alex Bewley, Ingmar Posner

Appearance changes due to weather and seasonal conditions represent a strong impediment to the robust implementation of machine learning systems in outdoor robotics.

Autonomous Driving Motion Planning +1

Incorporating Human Domain Knowledge into Large Scale Cost Function Learning

no code implementations13 Dec 2016 Markus Wulfmeier, Dushyant Rao, Ingmar Posner

Recent advances have shown the capability of Fully Convolutional Neural Networks (FCN) to model cost functions for motion planning in the context of learning driving preferences purely based on demonstration data from human drivers.

Motion Planning reinforcement-learning +1

Find Your Own Way: Weakly-Supervised Segmentation of Path Proposals for Urban Autonomy

no code implementations5 Oct 2016 Dan Barnes, Will Maddern, Ingmar Posner

We present a weakly-supervised approach to segmenting proposed drivable paths in images with the goal of autonomous driving in complex urban environments.

Autonomous Driving Segmentation +2

Deep Tracking on the Move: Learning to Track the World from a Moving Vehicle using Recurrent Neural Networks

no code implementations29 Sep 2016 Julie Dequaire, Dushyant Rao, Peter Ondruska, Dominic Wang, Ingmar Posner

This paper presents an end-to-end approach for tracking static and dynamic objects for an autonomous vehicle driving through crowded urban environments.

Watch This: Scalable Cost-Function Learning for Path Planning in Urban Environments

no code implementations8 Jul 2016 Markus Wulfmeier, Dominic Zeng Wang, Ingmar Posner

In this work, we present an approach to learn cost maps for driving in complex urban environments from a very large number of demonstrations of driving behaviour by human experts.

End-to-End Tracking and Semantic Segmentation Using Recurrent Neural Networks

no code implementations18 Apr 2016 Peter Ondruska, Julie Dequaire, Dominic Zeng Wang, Ingmar Posner

In this work we present a novel end-to-end framework for tracking and classifying a robot's surroundings in complex, dynamic and only partially observable real-world environments.

Classification General Classification +1

Deep Tracking: Seeing Beyond Seeing Using Recurrent Neural Networks

1 code implementation2 Feb 2016 Peter Ondruska, Ingmar Posner

This paper presents to the best of our knowledge the first end-to-end object tracking approach which directly maps from raw sensor input to object tracks in sensor space without requiring any feature engineering or system identification in the form of plant or sensor models.

Feature Engineering Object +1

Maximum Entropy Deep Inverse Reinforcement Learning

1 code implementation17 Jul 2015 Markus Wulfmeier, Peter Ondruska, Ingmar Posner

This paper presents a general framework for exploiting the representational capacity of neural networks to approximate complex, nonlinear reward functions in the context of solving the inverse reinforcement learning (IRL) problem.

reinforcement-learning Reinforcement Learning (RL)

Modelling Observation Correlations for Active Exploration and Robust Object Detection

no code implementations18 Jan 2014 Javier Velez, Garrett Hemann, Albert S. Huang, Ingmar Posner, Nicholas Roy

In particular, the performance of detection algorithms is commonly sensitive to the position of the sensor relative to the objects in the scene.

object-detection Robust Object Detection +1

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