IMLE Policy: Fast and Sample Efficient Visuomotor Policy Learning via Implicit Maximum Likelihood Estimation

Recent advances in imitation learning, particularly using generative modelling techniques like diffusion, have enabled policies to capture complex multi-modal action distributions. However, these methods often require large datasets and multiple inference steps for action generation, posing challenges in robotics where the cost for data collection is high and computation resources are limited. To address this, we introduce IMLE Policy, a novel behaviour cloning approach based on Implicit Maximum Likelihood Estimation (IMLE). IMLE Policy excels in low-data regimes, effectively learning from minimal demonstrations and requiring 38\% less data on average to match the performance of baseline methods in learning complex multi-modal behaviours. Its simple generator-based architecture enables single-step action generation, improving inference speed by 97.3\% compared to Diffusion Policy, while outperforming single-step Flow Matching. We validate our approach across diverse manipulation tasks in simulated and real-world environments, showcasing its ability to capture complex behaviours under data constraints. Videos and code are provided on our project page: https://imle-policy.github.io/.

Affordance-Centric Policy Learning: Sample Efficient and Generalisable Robot Policy Learning using Affordance-Centric Task Frames

Affordances are central to robotic manipulation, where most tasks can be simplified to interactions with task-specific regions on objects. By focusing on these key regions, we can abstract away task-irrelevant information, simplifying the learning process, and enhancing generalisation. In this paper, we propose an affordance-centric policy-learning approach that centres and appropriately \textit{orients} a \textit{task frame} on these affordance regions allowing us to achieve both \textbf{intra-category invariance} -- where policies can generalise across different instances within the same object category -- and \textbf{spatial invariance} -- which enables consistent performance regardless of object placement in the environment. We propose a method to leverage existing generalist large vision models to extract and track these affordance frames, and demonstrate that our approach can learn manipulation tasks using behaviour cloning from as little as 10 demonstrations, with equivalent generalisation to an image-based policy trained on 305 demonstrations. We provide video demonstrations on our project site: https://affordance-policy.github.io.

DiffusionNOCS: Managing Symmetry and Uncertainty in Sim2Real Multi-Modal Category-level Pose Estimation

This paper addresses the challenging problem of category-level pose estimation. Current state-of-the-art methods for this task face challenges when dealing with symmetric objects and when attempting to generalize to new environments solely through synthetic data training. In this work, we address these challenges by proposing a probabilistic model that relies on diffusion to estimate dense canonical maps crucial for recovering partial object shapes as well as establishing correspondences essential for pose estimation. Furthermore, we introduce critical components to enhance performance by leveraging the strength of the diffusion models with multi-modal input representations. We demonstrate the effectiveness of our method by testing it on a range of real datasets. Despite being trained solely on our generated synthetic data, our approach achieves state-of-the-art performance and unprecedented generalization qualities, outperforming baselines, even those specifically trained on the target domain.

Gravity-aware Grasp Generation with Implicit Grasp Mode Selection for Underactuated Hands

To overcome the mechanical limitation of parallel-jaw grippers, in this paper, we present a gravity-aware grasp generation that supports both precision grasp and power grasp of underactuated hands. We propose a novel approach to generate a large-scale dataset with a gravity-rejection score and experimentally confirm that the combination of that score and classical success/fail binary classification is powerful: the former encourages stable grasps, such as power grasps or grasping the center of mass, while the latter rejects invalid grasps, such as colliding with other objects or attempting to grasp parts that are too large for the gripper. We also propose a rotation representation that is continuous on SO(3) and considers the grasp's physical meaning. Our simulation and real robot evaluation experiments demonstrate significant improvements from the baseline works, especially for heavy objects.

GS-Pose: Category-Level Object Pose Estimation via Geometric and Semantic Correspondence

Category-level pose estimation is a challenging task with many potential applications in computer vision and robotics. Recently, deep-learning-based approaches have made great progress, but are typically hindered by the need for large datasets of either pose-labelled real images or carefully tuned photorealistic simulators. This can be avoided by using only geometry inputs such as depth images to reduce the domain-gap but these approaches suffer from a lack of semantic information, which can be vital in the pose estimation problem. To resolve this conflict, we propose to utilize both geometric and semantic features obtained from a pre-trained foundation model.Our approach projects 2D features from this foundation model into 3D for a single object model per category, and then performs matching against this for new single view observations of unseen object instances with a trained matching network. This requires significantly less data to train than prior methods since the semantic features are robust to object texture and appearance. We demonstrate this with a rich evaluation, showing improved performance over prior methods with a fraction of the data required.

Learning Fabric Manipulation in the Real World with Human Videos

Fabric manipulation is a long-standing challenge in robotics due to the enormous state space and complex dynamics. Learning approaches stand out as promising for this domain as they allow us to learn behaviours directly from data. Most prior methods however rely heavily on simulation, which is still limited by the large sim-to-real gap of deformable objects or rely on large datasets. A promising alternative is to learn fabric manipulation directly from watching humans perform the task. In this work, we explore how demonstrations for fabric manipulation tasks can be collected directly by humans, providing an extremely natural and fast data collection pipeline. Then, using only a handful of such demonstrations, we show how a pick-and-place policy can be learned and deployed on a real robot, without any robot data collection at all. We demonstrate our approach on a fabric folding task, showing that our policy can reliably reach folded states from crumpled initial configurations. Videos are available at: https://sites.google.com/view/foldingbyhand

Learning Arbitrary-Goal Fabric Folding with One Hour of Real Robot Experience

Manipulating deformable objects, such as fabric, is a long standing problem in robotics, with state estimation and control posing a significant challenge for traditional methods. In this paper, we show that it is possible to learn fabric folding skills in only an hour of self-supervised real robot experience, without human supervision or simulation. Our approach relies on fully convolutional networks and the manipulation of visual inputs to exploit learned features, allowing us to create an expressive goal-conditioned pick and place policy that can be trained efficiently with real world robot data only. Folding skills are learned with only a sparse reward function and thus do not require reward function engineering, merely an image of the goal configuration. We demonstrate our method on a set of towel-folding tasks, and show that our approach is able to discover sequential folding strategies, purely from trial-and-error. We achieve state-of-the-art results without the need for demonstrations or simulation, used in prior approaches. Videos available at: https://sites.google.com/view/learningtofold

Model-free vision-based shaping of deformable plastic materials

We address the problem of shaping deformable plastic materials using non-prehensile actions. Shaping plastic objects is challenging, since they are difficult to model and to track visually. We study this problem, by using kinetic sand, a plastic toy material which mimics the physical properties of wet sand. Inspired by a pilot study where humans shape kinetic sand, we define two types of actions: \textit{pushing} the material from the sides and \textit{tapping} from above. The chosen actions are executed with a robotic arm using image-based visual servoing. From the current and desired view of the material, we define states based on visual features such as the outer contour shape and the pixel luminosity values. These are mapped to actions, which are repeated iteratively to reduce the image error until convergence is reached. For pushing, we propose three methods for mapping the visual state to an action. These include heuristic methods and a neural network, trained from human actions. We show that it is possible to obtain simple shapes with the kinetic sand, without explicitly modeling the material. Our approach is limited in the types of shapes it can achieve. A richer set of action types and multi-step reasoning is needed to achieve more sophisticated shapes.

Evaluating task-agnostic exploration for fixed-batch learning of arbitrary future tasks

Deep reinforcement learning has been shown to solve challenging tasks where large amounts of training experience is available, usually obtained online while learning the task. Robotics is a significant potential application domain for many of these algorithms, but generating robot experience in the real world is expensive, especially when each task requires a lengthy online training procedure. Off-policy algorithms can in principle learn arbitrary tasks from a diverse enough fixed dataset. In this work, we evaluate popular exploration methods by generating robotics datasets for the purpose of learning to solve tasks completely offline without any further interaction in the real world. We present results on three popular continuous control tasks in simulation, as well as continuous control of a high-dimensional real robot arm. Code documenting all algorithms, experiments, and hyper-parameters is available at https://github.com/qutrobotlearning/batchlearning.

Mirroring to Build Trust in Digital Assistants

We describe experiments towards building a conversational digital assistant that considers the preferred conversational style of the user. In particular, these experiments are designed to measure whether users prefer and trust an assistant whose conversational style matches their own. To this end we conducted a user study where subjects interacted with a digital assistant that responded in a way that either matched their conversational style, or did not. Using self-reported personality attributes and subjects' feedback on the interactions, we built models that can reliably predict a user's preferred conversational style.