Inverse rendering aims to estimate physical attributes of a scene, e.g., reflectance, geometry, and lighting, from image(s). Inverse rendering has been studied primarily for single objects or with methods that solve for only one of the scene attributes. We propose the first learning-based approach that jointly estimates albedo, normals, and lighting of an indoor scene from a single image. Our key contribution is the Residual Appearance Renderer (RAR), which can be trained to synthesize complex appearance effects (e.g., inter-reflection, cast shadows, near-field illumination, and realistic shading), which would be neglected otherwise. This enables us to perform self-supervised learning on real data using a reconstruction loss, based on re-synthesizing the input image from the estimated components. We finetune with real data after pretraining with synthetic data. To this end, we use physically-based rendering to create a large-scale synthetic dataset, which is a significant improvement over prior datasets. Experimental results show that our approach outperforms state-of-the-art methods that estimate one or more scene attributes.
Self-Supervised learning from multimodal image and text data allows deep neural networks to learn powerful features with no need of human annotated data. Web and Social Media platforms provide a virtually unlimited amount of this multimodal data. In this work we propose to exploit this free available data to learn a multimodal image and text embedding, aiming to leverage the semantic knowledge learnt in the text domain and transfer it to a visual model for semantic image retrieval. We demonstrate that the proposed pipeline can learn from images with associated textwithout supervision and analyze the semantic structure of the learnt joint image and text embedding space. We perform a thorough analysis and performance comparison of five different state of the art text embeddings in three different benchmarks. We show that the embeddings learnt with Web and Social Media data have competitive performances over supervised methods in the text based image retrieval task, and we clearly outperform state of the art in the MIRFlickr dataset when training in the target data. Further, we demonstrate how semantic multimodal image retrieval can be performed using the learnt embeddings, going beyond classical instance-level retrieval problems. Finally, we present a new dataset, InstaCities1M, composed by Instagram images and their associated texts that can be used for fair comparison of image-text embeddings.
Operating directly from raw high dimensional sensory inputs like images is still a challenge for robotic control. Recently, Reinforcement Learning methods have been proposed to solve specific tasks end-to-end, from pixels to torques. However, these approaches assume the access to a specified reward which may require specialized instrumentation of the environment. Furthermore, the obtained policy and representations tend to be task specific and may not transfer well. In this work we investigate completely self-supervised learning of a general image embedding and control primitives, based on finding the shortest time to reach any state. We also introduce a new structure for the state-action value function that builds a connection between model-free and model-based methods, and improves the performance of the learning algorithm. We experimentally demonstrate these findings in three simulated robotic tasks.
We propose a self-supervised learning framework that uses unlabeled monocular video sequences to generate large-scale supervision for training a Visual Odometry (VO) frontend, a network which computes pointwise data associations across images. Our self-improving method enables a VO frontend to learn over time, unlike other VO and SLAM systems which require time-consuming hand-tuning or expensive data collection to adapt to new environments. Our proposed frontend operates on monocular images and consists of a single multi-task convolutional neural network which outputs 2D keypoints locations, keypoint descriptors, and a novel point stability score. We use the output of VO to create a self-supervised dataset of point correspondences to retrain the frontend. When trained using VO at scale on 2.5 million monocular images from ScanNet, the stability classifier automatically discovers a ranking for keypoints that are not likely to help in VO, such as t-junctions across depth discontinuities, features on shadows and highlights, and dynamic objects like people. The resulting frontend outperforms both traditional methods (SIFT, ORB, AKAZE) and deep learning methods (SuperPoint and LF-Net) in a 3D-to-2D pose estimation task on ScanNet.
Methods proposed in the literature towards continual deep learning typically operate in a task-based sequential learning setup. A sequence of tasks is learned, one at a time, with all data of current task available but not of previous or future tasks. Task boundaries and identities are known at all times. This setup, however, is rarely encountered in practical applications. Therefore we investigate how to transform continual learning to an online setup. We develop a system that keeps on learning over time in a streaming fashion, with data distributions gradually changing and without the notion of separate tasks. To this end, we build on the work on Memory Aware Synapses, and show how this method can be made online by providing a protocol to decide i) when to update the importance weights, ii) which data to use to update them, and iii) how to accumulate the importance weights at each update step. Experimental results show the validity of the approach in the context of two applications: (self-)supervised learning of a face recognition model by watching soap series and learning a robot to avoid collisions.
We introduce the first completely unsupervised correspondence learning approach for deformable 3D shapes. Key to our model is the understanding that natural deformations (such as changes in pose) approximately preserve the metric structure of the surface, yielding a natural criterion to drive the learning process toward distortion-minimizing predictions. On this basis, we overcome the need for annotated data and replace it by a purely geometric criterion. The resulting learning model is class-agnostic, and is able to leverage any type of deformable geometric data for the training phase. In contrast to existing supervised approaches which specialize on the class seen at training time, we demonstrate stronger generalization as well as applicability to a variety of challenging settings. We showcase our method on a wide selection of correspondence benchmarks, where we outperform other methods in terms of accuracy, generalization, and efficiency.
As 360{\deg} cameras become prevalent in many autonomous systems (e.g., self-driving cars and drones), efficient 360{\deg} perception becomes more and more important. We propose a novel self-supervised learning approach for predicting the omnidirectional depth and camera motion from a 360{\deg} video. In particular, starting from the SfMLearner, which is designed for cameras with normal field-of-view, we introduce three key features to process 360{\deg} images efficiently. Firstly, we convert each image from equirectangular projection to cubic projection in order to avoid image distortion. In each network layer, we use Cube Padding (CP), which pads intermediate features from adjacent faces, to avoid image boundaries. Secondly, we propose a novel "spherical" photometric consistency constraint on the whole viewing sphere. In this way, no pixel will be projected outside the image boundary which typically happens in images with normal field-of-view. Finally, rather than naively estimating six independent camera motions (i.e., naively applying SfM-Learner to each face on a cube), we propose a novel camera pose consistency loss to ensure the estimated camera motions reaching consensus. To train and evaluate our approach, we collect a new PanoSUNCG dataset containing a large amount of 360{\deg} videos with groundtruth depth and camera motion. Our approach achieves state-of-the-art depth prediction and camera motion estimation on PanoSUNCG with faster inference speed comparing to equirectangular. In real-world indoor videos, our approach can also achieve qualitatively reasonable depth prediction by acquiring model pre-trained on PanoSUNCG.
In this paper we present a self-supervised method for representation learning utilizing two different modalities. Based on the observation that cross-modal information has a high semantic meaning we propose a method to effectively exploit this signal. For our approach we utilize video data since it is available on a large scale and provides easily accessible modalities given by RGB and optical flow. We demonstrate state-of-the-art performance on highly contested action recognition datasets in the context of self-supervised learning. We show that our feature representation also transfers to other tasks and conduct extensive ablation studies to validate our core contributions. Code and model can be found at https://github.com/nawidsayed/Cross-and-Learn.
In this work, we introduce a novel algorithm for solving the textbook question answering (TQA) task which describes more realistic QA problems compared to other recent tasks. We mainly focus on two related issues with analysis of the TQA dataset. First, solving the TQA problems requires to comprehend multi-modal contexts in complicated input data. To tackle this issue of extracting knowledge features from long text lessons and merging them with visual features, we establish a context graph from texts and images, and propose a new module f-GCN based on graph convolutional networks (GCN). Second, scientific terms are not spread over the chapters and subjects are split in the TQA dataset. To overcome this so called "out-of-domain" issue, before learning QA problems, we introduce a novel self-supervised open-set learning process without any annotations. The experimental results show that our model significantly outperforms prior state-of-the-art methods. Moreover, ablation studies validate that both methods of incorporating f-GCN for extracting knowledge from multi-modal contexts and our newly proposed self-supervised learning process are effective for TQA problems.
Contact-rich manipulation tasks in unstructured environments often require both haptic and visual feedback. However, it is non-trivial to manually design a robot controller that combines modalities with very different characteristics. While deep reinforcement learning has shown success in learning control policies for high-dimensional inputs, these algorithms are generally intractable to deploy on real robots due to sample complexity. We use self-supervision to learn a compact and multimodal representation of our sensory inputs, which can then be used to improve the sample efficiency of our policy learning. We evaluate our method on a peg insertion task, generalizing over different geometry, configurations, and clearances, while being robust to external perturbations. Results for simulated and real robot experiments are presented.
Prediction is an appealing objective for self-supervised learning of behavioral skills, particularly for autonomous robots. However, effectively utilizing predictive models for control, especially with raw image inputs, poses a number of major challenges. How should the predictions be used? What happens when they are inaccurate? In this paper, we tackle these questions by proposing a method for learning robotic skills from raw image observations, using only autonomously collected experience. We show that even an imperfect model can complete complex tasks if it can continuously retry, but this requires the model to not lose track of the objective (e.g., the object of interest). To enable a robot to continuously retry a task, we devise a self-supervised algorithm for learning image registration, which can keep track of objects of interest for the duration of the trial. We demonstrate that this idea can be combined with a video-prediction based controller to enable complex behaviors to be learned from scratch using only raw visual inputs, including grasping, repositioning objects, and non-prehensile manipulation. Our real-world experiments demonstrate that a model trained with 160 robot hours of autonomously collected, unlabeled data is able to successfully perform complex manipulation tasks with a wide range of objects not seen during training.
In this work, we propose a self-supervised learning method for affine image registration on 3D medical images. Unlike optimisation-based methods, our affine image registration network (AIRNet) is designed to directly estimate the transformation parameters between two input images without using any metric, which represents the quality of the registration, as the optimising function. But since it is costly to manually identify the transformation parameters between any two images, we leverage the abundance of cheap unlabelled data to generate a synthetic dataset for the training of the model. Additionally, the structure of AIRNet enables us to learn the discriminative features of the images which are useful for registration purpose. Our proposed method was evaluated on magnetic resonance images of the axial view of human brain and compared with the performance of a conventional image registration method. Experiments demonstrate that our approach achieves better overall performance on registration of images from different patients and modalities with 100x speed-up in execution time.
Consensus maximisation learning can provide self-supervision when different views are available of the same data. The distributional hypothesis provides another form of useful self-supervision from adjacent sentences which are plentiful in large unlabelled corpora. Motivated by the observation that different learning architectures tend to emphasise different aspects of sentence meaning, we present a new self-supervised learning framework for learning sentence representations which minimises the disagreement between two views of the same sentence where one view encodes the sentence with a recurrent neural network (RNN), and the other view encodes the same sentence with a simple linear model. After learning, the individual views (networks) result in higher quality sentence representations than their single-view learnt counterparts (learnt using only the distributional hypothesis) as judged by performance on standard downstream tasks. An ensemble of both views provides even better generalisation on both supervised and unsupervised downstream tasks. Also, importantly the ensemble of views trained with consensus maximisation between the two different architectures performs better on downstream tasks than an analogous ensemble made from the single-view trained counterparts.
Word embeddings have been shown to benefit from ensambling several word embedding sources, often carried out using straightforward mathematical operations over the set of word vectors. More recently, self-supervised learning has been used to find a lower-dimensional representation, similar in size to the individual word embeddings within the ensemble. However, these methods do not use the available manually labeled datasets that are often used solely for the purpose of evaluation. We propose to reconstruct an ensemble of word embeddings as an auxiliary task that regularises a main task while both tasks share the learned meta-embedding layer. We carry out intrinsic evaluation (6 word similarity datasets and 3 analogy datasets) and extrinsic evaluation (4 downstream tasks). For intrinsic task evaluation, supervision comes from various labeled word similarity datasets. Our experimental results show that the performance is improved for all word similarity datasets when compared to self-supervised learning methods with a mean increase of $11.33$ in Spearman correlation. Specifically, the proposed method shows the best performance in 4 out of 6 of word similarity datasets when using a cosine reconstruction loss and Brier's word similarity loss. Moreover, improvements are also made when performing word meta-embedding reconstruction in sequence tagging and sentence meta-embedding for sentence classification.
We propose a self-supervised learning method to jointly reason about spatial and temporal context for video recognition. Recent self-supervised approaches have used spatial context [9, 34] as well as temporal coherency [32] but a combination of the two requires extensive preprocessing such as tracking objects through millions of video frames [59] or computing optical flow to determine frame regions with high motion [30]. We propose to combine spatial and temporal context in one self-supervised framework without any heavy preprocessing. We divide multiple video frames into grids of patches and train a network to solve jigsaw puzzles on these patches from multiple frames. So the network is trained to correctly identify the position of a patch within a video frame as well as the position of a patch over time. We also propose a novel permutation strategy that outperforms random permutations while significantly reducing computational and memory constraints. We use our trained network for transfer learning tasks such as video activity recognition and demonstrate the strength of our approach on two benchmark video action recognition datasets without using a single frame from these datasets for unsupervised pretraining of our proposed video jigsaw network.
We propose a self-supervised framework for learning facial attributes by simply watching videos of a human face speaking, laughing, and moving over time. To perform this task, we introduce a network, Facial Attributes-Net (FAb-Net), that is trained to embed multiple frames from the same video face-track into a common low-dimensional space. With this approach, we make three contributions: first, we show that the network can leverage information from multiple source frames by predicting confidence/attention masks for each frame; second, we demonstrate that using a curriculum learning regime improves the learned embedding; finally, we demonstrate that the network learns a meaningful face embedding that encodes information about head pose, facial landmarks and facial expression, i.e. facial attributes, without having been supervised with any labelled data. We are comparable or superior to state-of-the-art self-supervised methods on these tasks and approach the performance of supervised methods.
Self-supervised learning of convolutional neural networks can harness large amounts of cheap unlabeled data to train powerful feature representations. As surrogate task, we jointly address ordering of visual data in the spatial and temporal domain. The permutations of training samples, which are at the core of self-supervision by ordering, have so far been sampled randomly from a fixed preselected set. Based on deep reinforcement learning we propose a sampling policy that adapts to the state of the network, which is being trained. Therefore, new permutations are sampled according to their expected utility for updating the convolutional feature representation. Experimental evaluation on unsupervised and transfer learning tasks demonstrates competitive performance on standard benchmarks for image and video classification and nearest neighbor retrieval.
In this paper we present ActiveStereoNet, the first deep learning solution for active stereo systems. Due to the lack of ground truth, our method is fully self-supervised, yet it produces precise depth with a subpixel precision of $1/30th$ of a pixel; it does not suffer from the common over-smoothing issues; it preserves the edges; and it explicitly handles occlusions. We introduce a novel reconstruction loss that is more robust to noise and texture-less patches, and is invariant to illumination changes. The proposed loss is optimized using a window-based cost aggregation with an adaptive support weight scheme. This cost aggregation is edge-preserving and smooths the loss function, which is key to allow the network to reach compelling results. Finally we show how the task of predicting invalid regions, such as occlusions, can be trained end-to-end without ground-truth. This component is crucial to reduce blur and particularly improves predictions along depth discontinuities. Extensive quantitatively and qualitatively evaluations on real and synthetic data demonstrate state of the art results in many challenging scenes.
We propose a novel method for learning convolutional neural image representations without manual supervision. We use motion cues in the form of optical flow, to supervise representations of static images. The obvious approach of training a network to predict flow from a single image can be needlessly difficult due to intrinsic ambiguities in this prediction task. We instead propose a much simpler learning goal: embed pixels such that the similarity between their embeddings matches that between their optical flow vectors. At test time, the learned deep network can be used without access to video or flow information and transferred to tasks such as image classification, detection, and segmentation. Our method, which significantly simplifies previous attempts at using motion for self-supervision, achieves state-of-the-art results in self-supervision using motion cues, competitive results for self-supervision in general, and is overall state of the art in self-supervised pretraining for semantic image segmentation, as demonstrated on standard benchmarks.
The immense success of deep learning based methods in computer vision heavily relies on large scale training datasets. These richly annotated datasets help the network learn discriminative visual features. Collecting and annotating such datasets requires a tremendous amount of human effort and annotations are limited to popular set of classes. As an alternative, learning visual features by designing auxiliary tasks which make use of freely available self-supervision has become increasingly popular in the computer vision community. In this paper, we put forward an idea to take advantage of multi-modal context to provide self-supervision for the training of computer vision algorithms. We show that adequate visual features can be learned efficiently by training a CNN to predict the semantic textual context in which a particular image is more probable to appear as an illustration. More specifically we use popular text embedding techniques to provide the self-supervision for the training of deep CNN. Our experiments demonstrate state-of-the-art performance in image classification, object detection, and multi-modal retrieval compared to recent self-supervised or naturally-supervised approaches.