DyST: Towards Dynamic Neural Scene Representations on Real-World Videos

Visual understanding of the world goes beyond the semantics and flat structure of individual images. In this work, we aim to capture both the 3D structure and dynamics of real-world scenes from monocular real-world videos. Our Dynamic Scene Transformer (DyST) model leverages recent work in neural scene representation to learn a latent decomposition of monocular real-world videos into scene content, per-view scene dynamics, and camera pose. This separation is achieved through a novel co-training scheme on monocular videos and our new synthetic dataset DySO. DyST learns tangible latent representations for dynamic scenes that enable view generation with separate control over the camera and the content of the scene.

Object-Centric Learning for Real-World Videos by Predicting Temporal Feature Similarities

Unsupervised video-based object-centric learning is a promising avenue to learn structured representations from large, unlabeled video collections, but previous approaches have only managed to scale to real-world datasets in restricted domains. Recently, it was shown that the reconstruction of pre-trained self-supervised features leads to object-centric representations on unconstrained real-world image datasets. Building on this approach, we propose a novel way to use such pre-trained features in the form of a temporal feature similarity loss. This loss encodes temporal correlations between image patches and is a natural way to introduce a motion bias for object discovery. We demonstrate that this loss leads to state-of-the-art performance on the challenging synthetic MOVi datasets. When used in combination with the feature reconstruction loss, our model is the first object-centric video model that scales to unconstrained video datasets such as YouTube-VIS.

Bridging the Gap to Real-World Object-Centric Learning

Humans naturally decompose their environment into entities at the appropriate level of abstraction to act in the world. Allowing machine learning algorithms to derive this decomposition in an unsupervised way has become an important line of research. However, current methods are restricted to simulated data or require additional information in the form of motion or depth in order to successfully discover objects. In this work, we overcome this limitation by showing that reconstructing features from models trained in a self-supervised manner is a sufficient training signal for object-centric representations to arise in a fully unsupervised way. Our approach, DINOSAUR, significantly out-performs existing object-centric learning models on simulated data and is the first unsupervised object-centric model that scales to real world-datasets such as COCO and PASCAL VOC. DINOSAUR is conceptually simple and shows competitive performance compared to more involved pipelines from the computer vision literature.

On the Pitfalls of Heteroscedastic Uncertainty Estimation with Probabilistic Neural Networks

Capturing aleatoric uncertainty is a critical part of many machine learning systems. In deep learning, a common approach to this end is to train a neural network to estimate the parameters of a heteroscedastic Gaussian distribution by maximizing the logarithm of the likelihood function under the observed data. In this work, we examine this approach and identify potential hazards associated with the use of log-likelihood in conjunction with gradient-based optimizers. First, we present a synthetic example illustrating how this approach can lead to very poor but stable parameter estimates. Second, we identify the culprit to be the log-likelihood loss, along with certain conditions that exacerbate the issue. Third, we present an alternative formulation, termed $\beta$-NLL, in which each data point's contribution to the loss is weighted by the $\beta$-exponentiated variance estimate. We show that using an appropriate $\beta$ largely mitigates the issue in our illustrative example. Fourth, we evaluate this approach on a range of domains and tasks and show that it achieves considerable improvements and performs more robustly concerning hyperparameters, both in predictive RMSE and log-likelihood criteria.

Causal Influence Detection for Improving Efficiency in Reinforcement Learning

Many reinforcement learning (RL) environments consist of independent entities that interact sparsely. In such environments, RL agents have only limited influence over other entities in any particular situation. Our idea in this work is that learning can be efficiently guided by knowing when and what the agent can influence with its actions. To achieve this, we introduce a measure of situation-dependent causal influence based on conditional mutual information and show that it can reliably detect states of influence. We then propose several ways to integrate this measure into RL algorithms to improve exploration and off-policy learning. All modified algorithms show strong increases in data efficiency on robotic manipulation tasks.

Self-supervised Visual Reinforcement Learning with Object-centric Representations

Autonomous agents need large repertoires of skills to act reasonably on new tasks that they have not seen before. However, acquiring these skills using only a stream of high-dimensional, unstructured, and unlabeled observations is a tricky challenge for any autonomous agent. Previous methods have used variational autoencoders to encode a scene into a low-dimensional vector that can be used as a goal for an agent to discover new skills. Nevertheless, in compositional/multi-object environments it is difficult to disentangle all the factors of variation into such a fixed-length representation of the whole scene. We propose to use object-centric representations as a modular and structured observation space, which is learned with a compositional generative world model. We show that the structure in the representations in combination with goal-conditioned attention policies helps the autonomous agent to discover and learn useful skills. These skills can be further combined to address compositional tasks like the manipulation of several different objects.

NeurIPS 2019 Disentanglement Challenge: Improved Disentanglement through Learned Aggregation of Convolutional Feature Maps

This report to our stage 2 submission to the NeurIPS 2019 disentanglement challenge presents a simple image preprocessing method for learning disentangled latent factors. We propose to train a variational autoencoder on regionally aggregated feature maps obtained from networks pretrained on the ImageNet database, utilizing the implicit inductive bias contained in those features for disentanglement. This bias can be further enhanced by explicitly fine-tuning the feature maps on auxiliary tasks useful for the challenge, such as angle, position estimation, or color classification. Our approach achieved the 2nd place in stage 2 of the challenge. Code is available at https://github.com/mseitzer/neurips2019-disentanglement-challenge.

NeurIPS 2019 Disentanglement Challenge: Improved Disentanglement through Aggregated Convolutional Feature Maps

This report to our stage 1 submission to the NeurIPS 2019 disentanglement challenge presents a simple image preprocessing method for training VAEs leading to improved disentanglement compared to directly using the images. In particular, we propose to use regionally aggregated feature maps extracted from CNNs pretrained on ImageNet. Our method achieved the 2nd place in stage 1 of the challenge. Code is available at https://github.com/mseitzer/neurips2019-disentanglement-challenge.

Adversarial and Perceptual Refinement for Compressed Sensing MRI Reconstruction

Deep learning approaches have shown promising performance for compressed sensing-based Magnetic Resonance Imaging. While deep neural networks trained with mean squared error (MSE) loss functions can achieve high peak signal to noise ratio, the reconstructed images are often blurry and lack sharp details, especially for higher undersampling rates. Recently, adversarial and perceptual loss functions have been shown to achieve more visually appealing results. However, it remains an open question how to (1) optimally combine these loss functions with the MSE loss function and (2) evaluate such a perceptual enhancement. In this work, we propose a hybrid method, in which a visual refinement component is learnt on top of an MSE loss-based reconstruction network. In addition, we introduce a semantic interpretability score, measuring the visibility of the region of interest in both ground truth and reconstructed images, which allows us to objectively quantify the usefulness of the image quality for image post-processing and analysis. Applied on a large cardiac MRI dataset simulated with 8-fold undersampling, we demonstrate significant improvements ($p<0.01$) over the state-of-the-art in both a human observer study and the semantic interpretability score.