In this paper, we propose a novel pipeline that leverages language foundation models for temporal sequential pattern mining, such as for human mobility forecasting tasks. For example, in the task of predicting Place-of-Interest (POI) customer flows, typically the number of visits is extracted from historical logs, and only the numerical data are used to predict visitor flows. In this research, we perform the forecasting task directly on the natural language input that includes all kinds of information such as numerical values and contextual semantic information. Specific prompts are introduced to transform numerical temporal sequences into sentences so that existing language models can be directly applied. We design an AuxMobLCast pipeline for predicting the number of visitors in each POI, integrating an auxiliary POI category classification task with the encoder-decoder architecture. This research provides empirical evidence of the effectiveness of the proposed AuxMobLCast pipeline to discover sequential patterns in mobility forecasting tasks. The results, evaluated on three real-world datasets, demonstrate that pre-trained language foundation models also have good performance in forecasting temporal sequences. This study could provide visionary insights and lead to new research directions for predicting human mobility.
Decoders built on Gaussian processes (GPs) are enticing due to the marginalisation over the non-linear function space. Such models (also known as GP-LVMs) are often expensive and notoriously difficult to train in practice, but can be scaled using variational inference and inducing points. In this paper, we revisit active set approximations. We develop a new stochastic estimate of the log-marginal likelihood based on recently discovered links to cross-validation, and propose a computationally efficient approximation thereof. We demonstrate that the resulting stochastic active sets (SAS) approximation significantly improves the robustness of GP decoder training while reducing computational cost. The SAS-GP obtains more structure in the latent space, scales to many datapoints and learns better representations than variational autoencoders, which is rarely the case for GP decoders.
Multi-Document Scientific Summarization (MDSS) aims to produce coherent and concise summaries for clusters of topic-relevant scientific papers. This task requires precise understanding of paper content and accurate modeling of cross-paper relationships. Knowledge graphs convey compact and interpretable structured information for documents, which makes them ideal for content modeling and relationship modeling. In this paper, we present KGSum, an MDSS model centred on knowledge graphs during both the encoding and decoding process. Specifically, in the encoding process, two graph-based modules are proposed to incorporate knowledge graph information into paper encoding, while in the decoding process, we propose a two-stage decoder by first generating knowledge graph information of summary in the form of descriptive sentences, followed by generating the final summary. Empirical results show that the proposed architecture brings substantial improvements over baselines on the Multi-Xscience dataset.
Sparse representations of images are useful in many computer vision applications. Sparse coding with an $l_1$ penalty and a learned linear dictionary requires regularization of the dictionary to prevent a collapse in the $l_1$ norms of the codes. Typically, this regularization entails bounding the Euclidean norms of the dictionary's elements. In this work, we propose a novel sparse coding protocol which prevents a collapse in the codes without the need to regularize the decoder. Our method regularizes the codes directly so that each latent code component has variance greater than a fixed threshold over a set of sparse representations for a given set of inputs. Furthermore, we explore ways to effectively train sparse coding systems with multi-layer decoders since they can model more complex relationships than linear dictionaries. In our experiments with MNIST and natural image patches, we show that decoders learned with our approach have interpretable features both in the linear and multi-layer case. Moreover, we show that sparse autoencoders with multi-layer decoders trained using our variance regularization method produce higher quality reconstructions with sparser representations when compared to autoencoders with linear dictionaries. Additionally, sparse representations obtained with our variance regularization approach are useful in the downstream tasks of denoising and classification in the low-data regime.
We propose a two-stage approach for training a single NMT model to translate unseen languages both to and from English. For the first stage, we initialize an encoder-decoder model to pretrained XLM-R and RoBERTa weights, then perform multilingual fine-tuning on parallel data in 40 languages to English. We find this model can generalize to zero-shot translations on unseen languages. For the second stage, we leverage this generalization ability to generate synthetic parallel data from monolingual datasets, then bidirectionally train with successive rounds of back-translation. Our approach, which we EcXTra (English-centric Crosslingual (X) Transfer), is conceptually simple, only using a standard cross-entropy objective throughout. It is also data-driven, sequentially leveraging auxiliary parallel data and monolingual data. We evaluate unsupervised NMT results for 7 low-resource languages, and find that each round of back-translation training further refines bidirectional performance. Our final single EcXTra-trained model achieves competitive translation performance in all translation directions, notably establishing a new state-of-the-art for English-to-Kazakh (22.9 > 10.4 BLEU). Our code is available at https://github.com/manestay/EcXTra .
Position emission tomography (PET) is widely used in clinics and research due to its quantitative merits and high sensitivity, but suffers from low signal-to-noise ratio (SNR). Recently convolutional neural networks (CNNs) have been widely used to improve PET image quality. Though successful and efficient in local feature extraction, CNN cannot capture long-range dependencies well due to its limited receptive field. Global multi-head self-attention (MSA) is a popular approach to capture long-range information. However, the calculation of global MSA for 3D images has high computational costs. In this work, we proposed an efficient spatial and channel-wise encoder-decoder transformer, Spach Transformer, that can leverage spatial and channel information based on local and global MSAs. Experiments based on datasets of different PET tracers, i.e., $^{18}$F-FDG, $^{18}$F-ACBC, $^{18}$F-DCFPyL, and $^{68}$Ga-DOTATATE, were conducted to evaluate the proposed framework. Quantitative results show that the proposed Spach Transformer framework outperforms state-of-the-art deep learning architectures. Our codes are available at https://github.com/sijang/SpachTransformer
Video snapshot compressive imaging (SCI) captures multiple sequential video frames by a single measurement using the idea of computational imaging. The underlying principle is to modulate high-speed frames through different masks and these modulated frames are summed to a single measurement captured by a low-speed 2D sensor (dubbed optical encoder); following this, algorithms are employed to reconstruct the desired high-speed frames (dubbed software decoder) if needed. In this paper, we consider the reconstruction algorithm in video SCI, i.e., recovering a series of video frames from a compressed measurement. Specifically, we propose a Spatial-Temporal transFormer (STFormer) to exploit the correlation in both spatial and temporal domains. STFormer network is composed of a token generation block, a video reconstruction block, and these two blocks are connected by a series of STFormer blocks. Each STFormer block consists of a spatial self-attention branch, a temporal self-attention branch and the outputs of these two branches are integrated by a fusion network. Extensive results on both simulated and real data demonstrate the state-of-the-art performance of STFormer. The code and models are publicly available at https://github.com/ucaswangls/STFormer.git
Medical image segmentation assists in computer-aided diagnosis, surgeries, and treatment. Digitize tissue slide images are used to analyze and segment glands, nuclei, and other biomarkers which are further used in computer-aided medical applications. To this end, many researchers developed different neural networks to perform segmentation on histological images, mostly these networks are based on encoder-decoder architecture and also utilize complex attention modules or transformers. However, these networks are less accurate to capture relevant local and global features with accurate boundary detection at multiple scales, therefore, we proposed an Encoder-Decoder Network, Quick Attention Module and a Multi Loss Function (combination of Binary Cross Entropy (BCE) Loss, Focal Loss & Dice Loss). We evaluate the generalization capability of our proposed network on two publicly available datasets for medical image segmentation MoNuSeg and GlaS and outperform the state-of-the-art networks with 1.99% improvement on the MoNuSeg dataset and 7.15% improvement on the GlaS dataset. Implementation Code is available at this link: https://bit.ly/HistoSeg
This paper introduces a unified framework for video action segmentation via sequence to sequence (seq2seq) translation in a fully and timestamp supervised setup. In contrast to current state-of-the-art frame-level prediction methods, we view action segmentation as a seq2seq translation task, i.e., mapping a sequence of video frames to a sequence of action segments. Our proposed method involves a series of modifications and auxiliary loss functions on the standard Transformer seq2seq translation model to cope with long input sequences opposed to short output sequences and relatively few videos. We incorporate an auxiliary supervision signal for the encoder via a frame-wise loss and propose a separate alignment decoder for an implicit duration prediction. Finally, we extend our framework to the timestamp supervised setting via our proposed constrained k-medoids algorithm to generate pseudo-segmentations. Our proposed framework performs consistently on both fully and timestamp supervised settings, outperforming or competing state-of-the-art on several datasets. Our code is publicly available at https://github.com/boschresearch/UVAST.
Automatic medical question summarization can significantly help the system to understand consumer health questions and retrieve correct answers. The Seq2Seq model based on maximum likelihood estimation (MLE) has been applied in this task, which faces two general problems: the model can not capture well question focus and and the traditional MLE strategy lacks the ability to understand sentence-level semantics. To alleviate these problems, we propose a novel question focus-driven contrastive learning framework (QFCL). Specially, we propose an easy and effective approach to generate hard negative samples based on the question focus, and exploit contrastive learning at both encoder and decoder to obtain better sentence level representations. On three medical benchmark datasets, our proposed model achieves new state-of-the-art results, and obtains a performance gain of 5.33, 12.85 and 3.81 points over the baseline BART model on three datasets respectively. Further human judgement and detailed analysis prove that our QFCL model learns better sentence representations with the ability to distinguish different sentence meanings, and generates high-quality summaries by capturing question focus.
In large-scale retrieval, the lexicon-weighting paradigm, learning weighted sparse representations in vocabulary space, has shown promising results with high quality and low latency. Despite it deeply exploiting the lexicon-representing capability of pre-trained language models, a crucial gap remains between language modeling and lexicon-weighting retrieval -- the former preferring certain or low-entropy words whereas the latter favoring pivot or high-entropy words -- becoming the main barrier to lexicon-weighting performance for large-scale retrieval. To bridge this gap, we propose a brand-new pre-training framework, lexicon-bottlenecked masked autoencoder (LexMAE), to learn importance-aware lexicon representations. Essentially, we present a lexicon-bottlenecked module between a normal language modeling encoder and a weakened decoder, where a continuous bag-of-words bottleneck is constructed to learn a lexicon-importance distribution in an unsupervised fashion. The pre-trained LexMAE is readily transferred to the lexicon-weighting retrieval via fine-tuning. On the ad-hoc retrieval benchmark, MS-Marco, it achieves 42.6% MRR@10 with 45.8 QPS for the passage dataset and 44.4% MRR@100 with 134.8 QPS for the document dataset, by a CPU machine. And LexMAE shows state-of-the-art zero-shot transfer capability on BEIR benchmark with 12 datasets.
Multi-modal MR imaging is routinely used in clinical practice to diagnose and investigate brain tumors by providing rich complementary information. Previous multi-modal MRI segmentation methods usually perform modal fusion by concatenating multi-modal MRIs at an early/middle stage of the network, which hardly explores non-linear dependencies between modalities. In this work, we propose a novel Nested Modality-Aware Transformer (NestedFormer) to explicitly explore the intra-modality and inter-modality relationships of multi-modal MRIs for brain tumor segmentation. Built on the transformer-based multi-encoder and single-decoder structure, we perform nested multi-modal fusion for high-level representations of different modalities and apply modality-sensitive gating (MSG) at lower scales for more effective skip connections. Specifically, the multi-modal fusion is conducted in our proposed Nested Modality-aware Feature Aggregation (NMaFA) module, which enhances long-term dependencies within individual modalities via a tri-orientated spatial-attention transformer, and further complements key contextual information among modalities via a cross-modality attention transformer. Extensive experiments on BraTS2020 benchmark and a private meningiomas segmentation (MeniSeg) dataset show that the NestedFormer clearly outperforms the state-of-the-arts. The code is available at https://github.com/920232796/NestedFormer.
Anomalies in time-series provide insights of critical scenarios across a range of industries, from banking and aerospace to information technology, security, and medicine. However, identifying anomalies in time-series data is particularly challenging due to the imprecise definition of anomalies, the frequent absence of labels, and the enormously complex temporal correlations present in such data. The LSTM Autoencoder is an Encoder-Decoder scheme for Anomaly Detection based on Long Short Term Memory Networks that learns to reconstruct time-series behavior and then uses reconstruction error to identify abnormalities. We introduce the Denoising Architecture as a complement to this LSTM Encoder-Decoder model and investigate its effect on real-world as well as artificially generated datasets. We demonstrate that the proposed architecture increases both the accuracy and the training speed, thereby, making the LSTM Autoencoder more efficient for unsupervised anomaly detection tasks.
The intricacy of 3D surfaces often results cutting-edge point cloud denoising (PCD) models in surface degradation including remnant noise, wrongly-removed geometric details. Although using multi-scale patches to encode the geometry of a point has become the common wisdom in PCD, we find that simple aggregation of extracted multi-scale features can not adaptively utilize the appropriate scale information according to the geometric information around noisy points. It leads to surface degradation, especially for points close to edges and points on complex curved surfaces. We raise an intriguing question -- if employing multi-scale geometric perception information to guide the network to utilize multi-scale information, can eliminate the severe surface degradation problem? To answer it, we propose a Multi-offset Denoising Network (MODNet) customized for multi-scale patches. First, we extract the low-level feature of three scales patches by patch feature encoders. Second, a multi-scale perception module is designed to embed multi-scale geometric information for each scale feature and regress multi-scale weights to guide a multi-offset denoising displacement. Third, a multi-offset decoder regresses three scale offsets, which are guided by the multi-scale weights to predict the final displacement by weighting them adaptively. Experiments demonstrate that our method achieves new state-of-the-art performance on both synthetic and real-scanned datasets.
Automated abdominal multi-organ segmentation is a crucial yet challenging task in the computer-aided diagnosis of abdominal organ-related diseases. Although numerous deep learning models have achieved remarkable success in many medical image segmentation tasks, accurate segmentation of abdominal organs remains challenging, due to the varying sizes of abdominal organs and the ambiguous boundaries among them. In this paper, we propose a boundary-aware network (BA-Net) to segment abdominal organs on CT scans and MRI scans. This model contains a shared encoder, a boundary decoder, and a segmentation decoder. The multi-scale deep supervision strategy is adopted on both decoders, which can alleviate the issues caused by variable organ sizes. The boundary probability maps produced by the boundary decoder at each scale are used as attention to enhance the segmentation feature maps. We evaluated the BA-Net on the Abdominal Multi-Organ Segmentation (AMOS) Challenge dataset and achieved an average Dice score of 89.29$\%$ for multi-organ segmentation on CT scans and an average Dice score of 71.92$\%$ on MRI scans. The results demonstrate that BA-Net is superior to nnUNet on both segmentation tasks.
Purpose: Training auto-encoders on simulated signal evolution and inserting the decoder into the forward model improves reconstructions through more compact, Bloch-equation-based representations of signal in comparison to linear subspaces. Methods: Building on model-based nonlinear and linear subspace techniques that enable reconstruction of signal dynamics, we train auto-encoders on dictionaries of simulated signal evolution to learn more compact, non-linear, latent representations. The proposed Latent Signal Model framework inserts the decoder portion of the auto-encoder into the forward model and directly reconstructs the latent representation. Latent Signal Models essentially serve as a proxy for fast and feasible differentiation through the Bloch-equations used to simulate signal. This work performs experiments in the context of T2-shuffling, gradient echo EPTI, and MPRAGE-shuffling. We compare how efficiently auto-encoders represent signal evolution in comparison to linear subspaces. Simulation and in-vivo experiments then evaluate if reducing degrees of freedom by inserting the decoder into the forward model improves reconstructions in comparison to subspace constraints. Results: An auto-encoder with one real latent variable represents FSE, EPTI, and MPRAGE signal evolution as well as linear subspaces characterized by four basis vectors. In simulated/in-vivo T2-shuffling and in-vivo EPTI experiments, the proposed framework achieves consistent quantitative NRMSE and qualitative improvement over linear approaches. From qualitative evaluation, the proposed approach yields images with reduced blurring and noise amplification in MPRAGE shuffling experiments. Conclusion: Directly solving for non-linear latent representations of signal evolution improves time-resolved MRI reconstructions through reduced degrees of freedom.
Fine-tuning can be vulnerable to adversarial attacks. Existing works about black-box attacks on fine-tuned models (BAFT) are limited by strong assumptions. To fill the gap, we propose two novel BAFT settings, cross-domain and cross-domain cross-architecture BAFT, which only assume that (1) the target model for attacking is a fine-tuned model, and (2) the source domain data is known and accessible. To successfully attack fine-tuned models under both settings, we propose to first train an adversarial generator against the source model, which adopts an encoder-decoder architecture and maps a clean input to an adversarial example. Then we search in the low-dimensional latent space produced by the encoder of the adversarial generator. The search is conducted under the guidance of the surrogate gradient obtained from the source model. Experimental results on different domains and different network architectures demonstrate that the proposed attack method can effectively and efficiently attack the fine-tuned models.
Kidney structures segmentation is a crucial yet challenging task in the computer-aided diagnosis of surgery-based renal cancer. Although numerous deep learning models have achieved remarkable success in many medical image segmentation tasks, accurate segmentation of kidney structures on computed tomography angiography (CTA) images remains challenging, due to the variable sizes of kidney tumors and the ambiguous boundaries between kidney structures and their surroundings. In this paper, we propose a boundary-aware network (BA-Net) to segment kidneys, kidney tumors, arteries, and veins on CTA scans. This model contains a shared encoder, a boundary decoder, and a segmentation decoder. The multi-scale deep supervision strategy is adopted on both decoders, which can alleviate the issues caused by variable tumor sizes. The boundary probability maps produced by the boundary decoder at each scale are used as attention to enhance the segmentation feature maps. We evaluated the BA-Net on the Kidney PArsing (KiPA) Challenge dataset and achieved an average Dice score of 89.65$\%$ for kidney structure segmentation on CTA scans using 4-fold cross-validation. The results demonstrate the effectiveness of the BA-Net.
In this paper, we develop the Laplacian pyramid-like autoencoder (LPAE) by adding the Laplacian pyramid (LP) concept widely used to analyze images in Signal Processing. LPAE decomposes an image into the approximation image and the detail image in the encoder part and then tries to reconstruct the original image in the decoder part using the two components. We use LPAE for experiments on classifications and super-resolution areas. Using the detail image and the smaller-sized approximation image as inputs of a classification network, our LPAE makes the model lighter. Moreover, we show that the performance of the connected classification networks has remained substantially high. In a super-resolution area, we show that the decoder part gets a high-quality reconstruction image by setting to resemble the structure of LP. Consequently, LPAE improves the original results by combining the decoder part of the autoencoder and the super-resolution network.
Latent space exploration is a technique that discovers interpretable latent directions and manipulates latent codes to edit various attributes in images generated by generative adversarial networks (GANs). However, in previous work, spatial control is limited to simple transformations (e.g., translation and rotation), and it is laborious to identify appropriate latent directions and adjust their parameters. In this paper, we tackle the problem of editing the StyleGAN image layout by annotating the image directly. To do so, we propose an interactive framework for manipulating latent codes in accordance with the user inputs. In our framework, the user annotates a StyleGAN image with locations they want to move or not and specifies a movement direction by mouse dragging. From these user inputs and initial latent codes, our latent transformer based on a transformer encoder-decoder architecture estimates the output latent codes, which are fed to the StyleGAN generator to obtain a result image. To train our latent transformer, we utilize synthetic data and pseudo-user inputs generated by off-the-shelf StyleGAN and optical flow models, without manual supervision. Quantitative and qualitative evaluations demonstrate the effectiveness of our method over existing methods.