Adapting to the structure of data distributions (such as symmetry and transformation invariances) is an important challenge in machine learning. Invariances can be built into the learning process by architecture design, or by augmenting the dataset. Both require a priori knowledge about the exact nature of the symmetries. Absent this knowledge, practitioners resort to expensive and time-consuming tuning. To address this problem, we propose a new approach to learn distributions of augmentation transforms, in a new \emph{Transformed Risk Minimization} (TRM) framework. In addition to predictive models, we also optimize over transformations chosen from a hypothesis space. As an algorithmic framework, our TRM method is (1) efficient (jointly learns augmentations and models in a \emph{single training loop}), (2) modular (works with \emph{any} training algorithm), and (3) general (handles \emph{both discrete and continuous} augmentations). We theoretically compare TRM with standard risk minimization, and give a PAC-Bayes upper bound on its generalization error. We propose to optimize this bound over a rich augmentation space via a new parametrization over compositions of blocks, leading to the new \emph{Stochastic Compositional Augmentation Learning} (SCALE) algorithm. We compare SCALE experimentally with prior methods (Fast AutoAugment and Augerino) on CIFAR10/100, SVHN . Additionally, we show that SCALE can correctly learn certain symmetries in the data distribution (recovering rotations on rotated MNIST) and can also improve calibration of the learned model.
How can artificial agents learn to solve many diverse tasks in complex visual environments in the absence of any supervision? We decompose this question into two problems: discovering new goals and learning to reliably achieve them. We introduce Latent Explorer Achiever (LEXA), a unified solution to these that learns a world model from image inputs and uses it to train an explorer and an achiever policy from imagined rollouts. Unlike prior methods that explore by reaching previously visited states, the explorer plans to discover unseen surprising states through foresight, which are then used as diverse targets for the achiever to practice. After the unsupervised phase, LEXA solves tasks specified as goal images zero-shot without any additional learning. LEXA substantially outperforms previous approaches to unsupervised goal-reaching, both on prior benchmarks and on a new challenging benchmark with a total of 40 test tasks spanning across four standard robotic manipulation and locomotion domains. LEXA further achieves goals that require interacting with multiple objects in sequence. Finally, to demonstrate the scalability and generality of LEXA, we train a single general agent across four distinct environments. Code and videos at https://orybkin.github.io/lexa/
Robot learning holds the promise of learning policies that generalize broadly. However, such generalization requires sufficiently diverse datasets of the task of interest, which can be prohibitively expensive to collect. In other fields, such as computer vision, it is common to utilize shared, reusable datasets, such as ImageNet, to overcome this challenge, but this has proven difficult in robotics. In this paper, we ask: what would it take to enable practical data reuse in robotics for end-to-end skill learning? We hypothesize that the key is to use datasets with multiple tasks and multiple domains, such that a new user that wants to train their robot to perform a new task in a new domain can include this dataset in their training process and benefit from cross-task and cross-domain generalization. To evaluate this hypothesis, we collect a large multi-domain and multi-task dataset, with 7,200 demonstrations constituting 71 tasks across 10 environments, and empirically study how this data can improve the learning of new tasks in new environments. We find that jointly training with the proposed dataset and 50 demonstrations of a never-before-seen task in a new domain on average leads to a 2x improvement in success rate compared to using target domain data alone. We also find that data for only a few tasks in a new domain can bridge the domain gap and make it possible for a robot to perform a variety of prior tasks that were only seen in other domains. These results suggest that reusing diverse multi-task and multi-domain datasets, including our open-source dataset, may pave the way for broader robot generalization, eliminating the need to re-collect data for each new robot learning project.
We address the problem of estimating the shape of a person's head, defined as the geometry of the complete head surface, from a video taken with a single moving camera, and determining the alignment of the fitted 3D head for all video frames, irrespective of the person's pose. 3D head reconstructions commonly tend to focus on perfecting the face reconstruction, leaving the scalp to a statistical approximation. Our goal is to reconstruct the head model of each person to enable future mixed reality applications. To do this, we recover a dense 3D reconstruction and camera information via structure-from-motion and multi-view stereo. These are then used in a new two-stage fitting process to recover the 3D head shape by iteratively fitting a 3D morphable model of the head with the dense reconstruction in canonical space and fitting it to each person's head, using both traditional facial landmarks and scalp features extracted from the head's segmentation mask. Our approach recovers consistent geometry for varying head shapes, from videos taken by different people, with different smartphones, and in a variety of environments from living rooms to outdoor spaces.
This paper focuses on the problem of 3D human reconstruction from 2D evidence. Although this is an inherently ambiguous problem, the majority of recent works avoid the uncertainty modeling and typically regress a single estimate for a given input. In contrast to that, in this work, we propose to embrace the reconstruction ambiguity and we recast the problem as learning a mapping from the input to a distribution of plausible 3D poses. Our approach is based on the normalizing flows model and offers a series of advantages. For conventional applications, where a single 3D estimate is required, our formulation allows for efficient mode computation. Using the mode leads to performance that is comparable with the state of the art among deterministic unimodal regression models. Simultaneously, since we have access to the likelihood of each sample, we demonstrate that our model is useful in a series of downstream tasks, where we leverage the probabilistic nature of the prediction as a tool for more accurate estimation. These tasks include reconstruction from multiple uncalibrated views, as well as human model fitting, where our model acts as a powerful image-based prior for mesh recovery. Our results validate the importance of probabilistic modeling, and indicate state-of-the-art performance across a variety of settings. Code and models are available at: https://www.seas.upenn.edu/~nkolot/projects/prohmr.
We consider the problem of object goal navigation in unseen environments. In our view, solving this problem requires learning of contextual semantic priors, a challenging endeavour given the spatial and semantic variability of indoor environments. Current methods learn to implicitly encode these priors through goal-oriented navigation policy functions operating on spatial representations that are limited to the agent's observable areas. In this work, we propose a novel framework that actively learns to generate semantic maps outside the field of view of the agent and leverages the uncertainty over the semantic classes in the unobserved areas to decide on long term goals. We demonstrate that through this spatial prediction strategy, we are able to learn semantic priors in scenes that can be leveraged in unknown environments. Additionally, we show how different objectives can be defined by balancing exploration with exploitation during searching for semantic targets. Our method is validated in the visually realistic environments offered by the Matterport3D dataset and show state of the art results on the object goal navigation task.
The ability to plan into the future while utilizing only raw high-dimensional observations, such as images, can provide autonomous agents with broad capabilities. Visual model-based reinforcement learning (RL) methods that plan future actions directly have shown impressive results on tasks that require only short-horizon reasoning, however, these methods struggle on temporally extended tasks. We argue that it is easier to solve long-horizon tasks by planning sequences of states rather than just actions, as the effects of actions greatly compound over time and are harder to optimize. To achieve this, we draw on the idea of collocation, which has shown good results on long-horizon tasks in optimal control literature, and adapt it to the image-based setting by utilizing learned latent state space models. The resulting latent collocation method (LatCo) optimizes trajectories of latent states, which improves over previously proposed shooting methods for visual model-based RL on tasks with sparse rewards and long-term goals. Videos and code at https://orybkin.github.io/latco/.
Animals are diverse in shape, but building a deformable shape model for a new species is not always possible due to the lack of 3D data. We present a method to capture new species using an articulated template and images of that species. In this work, we focus mainly on birds. Although birds represent almost twice the number of species as mammals, no accurate shape model is available. To capture a novel species, we first fit the articulated template to each training sample. By disentangling pose and shape, we learn a shape space that captures variation both among species and within each species from image evidence. We learn models of multiple species from the CUB dataset, and contribute new species-specific and multi-species shape models that are useful for downstream reconstruction tasks. Using a low-dimensional embedding, we show that our learned 3D shape space better reflects the phylogenetic relationships among birds than learned perceptual features.
In order to interact with the world, agents must be able to predict the results of the world's dynamics. A natural approach to learn about these dynamics is through video prediction, as cameras are ubiquitous and powerful sensors. Direct pixel-to-pixel video prediction is difficult, does not take advantage of known priors, and does not provide an easy interface to utilize the learned dynamics. Object-centric video prediction offers a solution to these problems by taking advantage of the simple prior that the world is made of objects and by providing a more natural interface for control. However, existing object-centric video prediction pipelines require dense object annotations in training video sequences. In this work, we present Object-centric Prediction without Annotation (OPA), an object-centric video prediction method that takes advantage of priors from powerful computer vision models. We validate our method on a dataset comprised of video sequences of stacked objects falling, and demonstrate how to adapt a perception model in an environment through end-to-end video prediction training.
We propose a framework for deformable linear object prediction. Prediction of deformable objects (e.g., rope) is challenging due to their non-linear dynamics and infinite-dimensional configuration spaces. By mapping the dynamics from a non-linear space to a linear space, we can use the good properties of linear dynamics for easier learning and more efficient prediction. We learn a locally linear, action-conditioned dynamics model that can be used to predict future latent states. Then, we decode the predicted latent state into the predicted state. We also apply a sampling-based optimization algorithm to select the optimal control action. We empirically demonstrate that our approach can predict the rope state accurately up to ten steps into the future and that our algorithm can find the optimal action given an initial state and a goal state.