We introduce $\textit{InExtremIS}$, a weakly supervised 3D approach to train a deep image segmentation network using particularly weak train-time annotations: only 6 extreme clicks at the boundary of the objects of interest. Our fully-automatic method is trained end-to-end and does not require any test-time annotations. From the extreme points, 3D bounding boxes are extracted around objects of interest. Then, deep geodesics connecting extreme points are generated to increase the amount of "annotated" voxels within the bounding boxes. Finally, a weakly supervised regularised loss derived from a Conditional Random Field formulation is used to encourage prediction consistency over homogeneous regions. Extensive experiments are performed on a large open dataset for Vestibular Schwannoma segmentation. $\textit{InExtremIS}$ obtained competitive performance, approaching full supervision and outperforming significantly other weakly supervised techniques based on bounding boxes. Moreover, given a fixed annotation time budget, $\textit{InExtremIS}$ outperforms full supervision. Our code and data are available online.
Machine learning models based on sensitive data in the real-world promise advances in areas ranging from medical screening to disease outbreaks, agriculture, industry, defense science, and more. In many applications, learning participant communication rounds benefit from collecting their own private data sets, teaching detailed machine learning models on the real data, and sharing the benefits of using these models. Due to existing privacy and security concerns, most people avoid sensitive data sharing for training. Without each user demonstrating their local data to a central server, Federated Learning allows various parties to train a machine learning algorithm on their shared data jointly. This method of collective privacy learning results in the expense of important communication during training. Most large-scale machine-learning applications require decentralized learning based on data sets generated on various devices and places. Such datasets represent an essential obstacle to decentralized learning, as their diverse contexts contribute to significant differences in the delivery of data across devices and locations. Researchers have proposed several ways to achieve data privacy in Federated Learning systems. However, there are still challenges with homogeneous local data. This research approach is to select nodes (users) to share their data in Federated Learning for independent data-based equilibrium to improve accuracy, reduce training time, and increase convergence. Therefore, this research presents a combined Deep-QReinforcement Learning Ensemble based on Spectral Clustering called DQRE-SCnet to choose a subset of devices in each communication round. Based on the results, it has been displayed that it is possible to decrease the number of communication rounds needed in Federated Learning.
Boolean matching is significant to digital integrated circuits design. An exhaustive method for Boolean matching is computationally expensive even for functions with only a few variables, because the time complexity of such an algorithm for an n-variable Boolean function is $O(2^{n+1}n!)$. Sensitivity is an important characteristic and a measure of the complexity of Boolean functions. It has been used in analysis of the complexity of algorithms in different fields. This measure could be regarded as a signature of Boolean functions and has great potential to help reduce the search space of Boolean matching. In this paper, we introduce Boolean sensitivity into Boolean matching and design several sensitivity-related signatures to enhance fast Boolean matching. First, we propose some new signatures that relate sensitivity to Boolean equivalence. Then, we prove that these signatures are prerequisites for Boolean matching, which we can use to reduce the search space of the matching problem. Besides, we develop a fast sensitivity calculation method to compute and compare these signatures of two Boolean functions. Compared with the traditional cofactor and symmetric detection methods, sensitivity is a series of signatures of another dimension. We also show that sensitivity can be easily integrated into traditional methods and distinguish the mismatched Boolean functions faster. To the best of our knowledge, this is the first work that introduces sensitivity to Boolean matching. The experimental results show that sensitivity-related signatures we proposed in this paper can reduce the search space to a very large extent, and perform up to 3x speedup over the state-of-the-art Boolean matching methods.
Vision transformers (ViTs) have pushed the state-of-the-art for various visual recognition tasks by patch-wise image tokenization followed by self-attention. However, the employment of self-attention modules results in a quadratic complexity in both computation and memory usage. Various attempts on approximating the self-attention computation with linear complexity have been made in Natural Language Processing. However, an in-depth analysis in this work shows that they are either theoretically flawed or empirically ineffective for visual recognition. We further identify that their limitations are rooted in keeping the softmax self-attention during approximations. Specifically, conventional self-attention is computed by normalizing the scaled dot-product between token feature vectors. Keeping this softmax operation challenges any subsequent linearization efforts. Based on this insight, for the first time, a softmax-free transformer or SOFT is proposed. To remove softmax in self-attention, Gaussian kernel function is used to replace the dot-product similarity without further normalization. This enables a full self-attention matrix to be approximated via a low-rank matrix decomposition. The robustness of the approximation is achieved by calculating its Moore-Penrose inverse using a Newton-Raphson method. Extensive experiments on ImageNet show that our SOFT significantly improves the computational efficiency of existing ViT variants. Crucially, with a linear complexity, much longer token sequences are permitted in SOFT, resulting in superior trade-off between accuracy and complexity.
The goal of this work is to efficiently identify visually similar patterns from a pair of images, e.g. identifying an artwork detail copied between an engraving and an oil painting, or matching a night-time photograph with its daytime counterpart. Lack of training data is a key challenge for this co-segmentation task. We present a simple yet surprisingly effective approach to overcome this difficulty: we generate synthetic training pairs by selecting object segments in an image and copy-pasting them into another image. We then learn to predict the repeated object masks. We find that it is crucial to predict the correspondences as an auxiliary task and to use Poisson blending and style transfer on the training pairs to generalize on real data. We analyse results with two deep architectures relevant to our joint image analysis task: a transformer-based architecture and Sparse Nc-Net, a recent network designed to predict coarse correspondences using 4D convolutions. We show our approach provides clear improvements for artwork details retrieval on the Brueghel dataset and achieves competitive performance on two place recognition benchmarks, Tokyo247 and Pitts30K. We then demonstrate the potential of our approach by performing object discovery on the Internet object discovery dataset and the Brueghel dataset. Our code and data are available at http://imagine.enpc.fr/~shenx/SegSwap/.
Dictionary learning aims to find a dictionary under which the training data can be sparsely represented, and it is usually achieved by iteratively applying two stages: sparse coding and dictionary update. Typical methods for dictionary update focuses on refining both dictionary atoms and their corresponding sparse coefficients by using the sparsity patterns obtained from sparse coding stage, and hence it is a non-convex bilinear inverse problem. In this paper, we propose a Rank-One Matrix Decomposition (ROMD) algorithm to recast this challenge into a convex problem by resolving these two variables into a set of rank-one matrices. Different from methods in the literature, ROMD updates the whole dictionary at a time using convex programming. The advantages hence include both convergence guarantees for dictionary update and faster convergence of the whole dictionary learning. The performance of ROMD is compared with other benchmark dictionary learning algorithms. The results show the improvement of ROMD in recovery accuracy, especially in the cases of high sparsity level and fewer observation data.
This paper describes a novel approach in human robot interaction driven by ergonomics. With a clear focus on optimising ergonomics, the approach proposed here continuously observes a human user's posture and by invoking appropriate cooperative robot movements, the user's posture is, whenever required, brought back to an ergonomic optimum. Effectively, the new protocol optimises the human-robot relative position and orientation as a function of human ergonomics. An RGB-D camera is used to calculate and monitor human joint angles in real-time and to determine the current ergonomics state. A total of 6 main causes of low ergonomic states are identified, leading to 6 universal robot responses to allow the human to return to an optimal ergonomics state. The algorithmic framework identifies these 6 causes and controls the cooperating robot to always adapt the environment (e.g. change the pose of the workpiece) in a way that is ergonomically most comfortable for the interacting user. Hence, human-robot interaction is continuously re-evaluated optimizing ergonomics states. The approach is validated through an experimental study, based on established ergonomic methods and their adaptation for real-time application. The study confirms improved ergonomics using the new approach.
Autonomous systems use extensively learning-enabled components such as deep neural networks (DNNs) for prediction and decision making. In this paper, we utilize a feedback loop between learning-enabled components used for classification and the sensors of an autonomous system in order to improve the confidence of the predictions. We design a classifier using Inductive Conformal Prediction (ICP) based on a triplet network architecture in order to learn representations that can be used to quantify the similarity between test and training examples. The method allows computing confident set predictions with an error rate predefined using a selected significance level. A feedback loop that queries the sensors for a new input is used to further refine the predictions and increase the classification accuracy. The method is computationally efficient, scalable to high-dimensional inputs, and can be executed in a feedback loop with the system in real-time. The approach is evaluated using a traffic sign recognition dataset and the results show that the error rate is reduced.
One of the main challenges in reinforcement learning is solving tasks with sparse reward. We show that the difficulty of discovering a distant rewarding state in an MDP is bounded by the expected cover time of a random walk over the graph induced by the MDP's transition dynamics. We therefore propose to accelerate exploration by constructing options that minimize cover time. The proposed algorithm finds an option which provably diminishes the expected number of steps to visit every state in the state space by a uniform random walk. We show empirically that the proposed algorithm improves the learning time in several domains with sparse rewards.
Video anomaly detection is often seen as one-class classification (OCC) problem due to the limited availability of anomaly examples. Typically, to tackle this problem, an autoencoder (AE) is trained to reconstruct the input with training set consisting only of normal data. At test time, the AE is then expected to well reconstruct the normal data while poorly reconstructing the anomalous data. However, several studies have shown that, even with only normal data training, AEs can often start reconstructing anomalies as well which depletes the anomaly detection performance. To mitigate this problem, we propose a novel methodology to train AEs with the objective of reconstructing only normal data, regardless of the input (i.e., normal or abnormal). Since no real anomalies are available in the OCC settings, the training is assisted by pseudo anomalies that are generated by manipulating normal data to simulate the out-of-normal-data distribution. We additionally propose two ways to generate pseudo anomalies: patch and skip frame based. Extensive experiments on three challenging video anomaly datasets demonstrate the effectiveness of our method in improving conventional AEs, achieving state-of-the-art performance.