The task of Chinese Spelling Check (CSC) is aiming to detect and correct spelling errors that can be found in the text. While manually annotating a high-quality dataset is expensive and time-consuming, thus the scale of the training dataset is usually very small (e.g., SIGHAN15 only contains 2339 samples for training), therefore supervised-learning based models usually suffer the data sparsity limitation and over-fitting issue, especially in the era of big language models. In this paper, we are dedicated to investigating the \textbf{unsupervised} paradigm to address the CSC problem and we propose a framework named \textbf{uChecker} to conduct unsupervised spelling error detection and correction. Masked pretrained language models such as BERT are introduced as the backbone model considering their powerful language diagnosis capability. Benefiting from the various and flexible MASKing operations, we propose a Confusionset-guided masking strategy to fine-train the masked language model to further improve the performance of unsupervised detection and correction. Experimental results on standard datasets demonstrate the effectiveness of our proposed model uChecker in terms of character-level and sentence-level Accuracy, Precision, Recall, and F1-Measure on tasks of spelling error detection and correction respectively.
The central object of this PhD thesis is known under different names in the fields of computer science and statistical mechanics. In computer science, it is called the Maximum Cut problem, one of the famous twenty-one Karp's original NP-hard problems, while the same object from Physics is called the Ising Spin Glass model. This model of a rich structure often appears as a reduction or reformulation of real-world problems from computer science, physics and engineering. However, solving this model exactly (finding the maximal cut or the ground state) is likely to stay an intractable problem (unless $\textit{P} = \textit{NP}$) and requires the development of ad-hoc heuristics for every particular family of instances. One of the bright and beautiful connections between discrete and continuous optimization is a Semidefinite Programming-based rounding scheme for Maximum Cut. This procedure allows us to find a provably near-optimal solution; moreover, this method is conjectured to be the best possible in polynomial time. In the first two chapters of this thesis, we investigate local non-convex heuristics intended to improve the rounding scheme. In the last chapter of this thesis, we make one step further and aim to control the solution of the problem we wanted to solve in previous chapters. We formulate a bi-level optimization problem over the Ising model where we want to tweak the interactions as little as possible so that the ground state of the resulting Ising model satisfies the desired criteria. This kind of problem arises in pandemic modeling. We show that when the interactions are non-negative, our bi-level optimization is solvable in polynomial time using convex programming.
The problem of reducing a Hidden Markov Model (HMM) to a one of smaller dimension that exactly reproduces the same marginals is tackled by using a system-theoretic approach, adapted to HMMs by leveraging on a suitable algebraic representation of probability spaces. We propose two algorithms that return coarse-grained equivalent HMMs obtained by stochastic projection operators: the first returns models that reproduce the single-time distribution of a given output process, while in the second the full (multi-time) distribution is preserved. The reduction method exploits not only the structure of the observed output, but also its initial condition, whenever the latter is known or belongs to a given subclass. Optimal algorithms are derived for a class of HMM, namely observable ones. In the general case, we propose algorithms that have produced minimal models for all the examples we analyzed, and conjecture their optimality.
Abdominal organ segmentation has many important clinical applications, such as organ quantification, surgical planning, and disease diagnosis. However, manually annotating organs from CT scans is time-consuming and labor-intensive. Semi-supervised learning has shown the potential to alleviate this challenge by learning from a large set of unlabeled images and limited labeled samples. In this work, we follow the self-training strategy and employ a hybrid architecture (PHTrans) with CNN and Transformer for both teacher and student models to generate precise pseudo-labels. Afterward, we introduce them with label data together into a two-stage segmentation framework with lightweight PHTrans for training to improve the performance and generalization ability of the model while remaining efficient. Experiments on the validation set of FLARE2022 demonstrate that our method achieves excellent segmentation performance as well as fast and low-resource model inference. The average DSC and HSD are 0.8956 and 0.9316, respectively. Under our development environments, the average inference time is 18.62 s, the average maximum GPU memory is 1995.04 MB, and the area under the GPU memory-time curve and the average area under the CPU utilization-time curve are 23196.84 and 319.67.
Unmanned aerial vehicles (UAVs) are widely applied for purposes of inspection, search, and rescue operations by the virtue of low-cost, large-coverage, real-time, and high-resolution data acquisition capacities. Massive volumes of aerial videos are produced in these processes, in which normal events often account for an overwhelming proportion. It is extremely difficult to localize and extract abnormal events containing potentially valuable information from long video streams manually. Therefore, we are dedicated to developing anomaly detection methods to solve this issue. In this paper, we create a new dataset, named DroneAnomaly, for anomaly detection in aerial videos. This dataset provides 37 training video sequences and 22 testing video sequences from 7 different realistic scenes with various anomalous events. There are 87,488 color video frames (51,635 for training and 35,853 for testing) with the size of $640 \times 640$ at 30 frames per second. Based on this dataset, we evaluate existing methods and offer a benchmark for this task. Furthermore, we present a new baseline model, ANomaly Detection with Transformers (ANDT), which treats consecutive video frames as a sequence of tubelets, utilizes a Transformer encoder to learn feature representations from the sequence, and leverages a decoder to predict the next frame. Our network models normality in the training phase and identifies an event with unpredictable temporal dynamics as an anomaly in the test phase. Moreover, To comprehensively evaluate the performance of our proposed method, we use not only our Drone-Anomaly dataset but also another dataset. We will make our dataset and code publicly available. A demo video is available at https://youtu.be/ancczYryOBY. We make our dataset and code publicly available .
Benefiting from the vigorous development of deep learning, many CNN-based image super-resolution methods have emerged and achieved better results than traditional algorithms. However, it is difficult for most algorithms to adaptively adjust the spatial region and channel features at the same time, let alone the information exchange between them. In addition, the exchange of information between attention modules is even less visible to researchers. To solve these problems, we put forward a lightweight spatial-channel adaptive coordination of multilevel refinement enhancement networks(MREN). Specifically, we construct a space-channel adaptive coordination block, which enables the network to learn the spatial region and channel feature information of interest under different receptive fields. In addition, the information of the corresponding feature processing level between the spatial part and the channel part is exchanged with the help of jump connection to achieve the coordination between the two. We establish a communication bridge between attention modules through a simple linear combination operation, so as to more accurately and continuously guide the network to pay attention to the information of interest. Extensive experiments on several standard test sets have shown that our MREN achieves superior performance over other advanced algorithms with a very small number of parameters and very low computational complexity.
In this paper we address the problem of neural face reenactment, where, given a pair of a source and a target facial image, we need to transfer the target's pose (defined as the head pose and its facial expressions) to the source image, by preserving at the same time the source's identity characteristics (e.g., facial shape, hair style, etc), even in the challenging case where the source and the target faces belong to different identities. In doing so, we address some of the limitations of the state-of-the-art works, namely, a) that they depend on paired training data (i.e., source and target faces have the same identity), b) that they rely on labeled data during inference, and c) that they do not preserve identity in large head pose changes. More specifically, we propose a framework that, using unpaired randomly generated facial images, learns to disentangle the identity characteristics of the face from its pose by incorporating the recently introduced style space $\mathcal{S}$ of StyleGAN2, a latent representation space that exhibits remarkable disentanglement properties. By capitalizing on this, we learn to successfully mix a pair of source and target style codes using supervision from a 3D model. The resulting latent code, that is subsequently used for reenactment, consists of latent units corresponding to the facial pose of the target only and of units corresponding to the identity of the source only, leading to notable improvement in the reenactment performance compared to recent state-of-the-art methods. In comparison to state of the art, we quantitatively and qualitatively show that the proposed method produces higher quality results even on extreme pose variations. Finally, we report results on real images by first embedding them on the latent space of the pretrained generator. We make the code and pretrained models publicly available at: https://github.com/StelaBou/StyleMask
In this paper, we present a framework for reading analog clocks in natural images or videos. Specifically, we make the following contributions: First, we create a scalable pipeline for generating synthetic clocks, significantly reducing the requirements for the labour-intensive annotations; Second, we introduce a clock recognition architecture based on spatial transformer networks (STN), which is trained end-to-end for clock alignment and recognition. We show that the model trained on the proposed synthetic dataset generalises towards real clocks with good accuracy, advocating a Sim2Real training regime; Third, to further reduce the gap between simulation and real data, we leverage the special property of time, i.e. uniformity, to generate reliable pseudo-labels on real unlabelled clock videos, and show that training on these videos offers further improvements while still requiring zero manual annotations. Lastly, we introduce three benchmark datasets based on COCO, Open Images, and The Clock movie, totalling 4,472 images with clocks, with full annotations for time, accurate to the minute.
The real-world deployment of fully autonomous mobile robots depends on a robust SLAM (Simultaneous Localization and Mapping) system, capable of handling dynamic environments, where objects are moving in front of the robot, and changing environments, where objects are moved or replaced after the robot has already mapped the scene. This paper presents Changing-SLAM, a method for robust Visual SLAM in both dynamic and changing environments. This is achieved by using a Bayesian filter combined with a long-term data association algorithm. Also, it employs an efficient algorithm for dynamic keypoints filtering based on object detection that correctly identify features inside the bounding box that are not dynamic, preventing a depletion of features that could cause lost tracks. Furthermore, a new dataset was developed with RGB-D data especially designed for the evaluation of changing environments on an object level, called PUC-USP dataset. Six sequences were created using a mobile robot, an RGB-D camera and a motion capture system. The sequences were designed to capture different scenarios that could lead to a tracking failure or a map corruption. To the best of our knowledge, Changing-SLAM is the first Visual SLAM system that is robust to both dynamic and changing environments, not assuming a given camera pose or a known map, being also able to operate in real time. The proposed method was evaluated using benchmark datasets and compared with other state-of-the-art methods, proving to be highly accurate.
Particle smoothers are SMC (Sequential Monte Carlo) algorithms designed to approximate the joint distribution of the states given observations from a state-space model. We propose dSMC (de-Sequentialized Monte Carlo), a new particle smoother that is able to process $T$ observations in $\mathcal{O}(\log T)$ time on parallel architecture. This compares favourably with standard particle smoothers, the complexity of which is linear in $T$. We derive $\mathcal{L}_p$ convergence results for dSMC, with an explicit upper bound, polynomial in $T$. We then discuss how to reduce the variance of the smoothing estimates computed by dSMC by (i) designing good proposal distributions for sampling the particles at the initialization of the algorithm, as well as by (ii) using lazy resampling to increase the number of particles used in dSMC. Finally, we design a particle Gibbs sampler based on dSMC, which is able to perform parameter inference in a state-space model at a $\mathcal{O}(\log(T))$ cost on parallel hardware.