Since satellite systems are playing an increasingly important role in our civilization, their security and privacy weaknesses are more and more concerned. For example, prior work demonstrates that the communication channel between maritime VSAT and ground segment can be eavesdropped on using consumer-grade equipment. The stream decoder GSExtract developed in this prior work performs well for most packets but shows incapacity for corrupted streams. We discovered that such stream corruption commonly exists in not only Europe and North Atlantic areas but also Asian areas. In our experiment, using GSExtract, we are only able to decode 2.1\% satellite streams we eavesdropped on in Asia. Therefore, in this work, we propose to use a contrastive learning technique with data augmentation to decode and recover such highly corrupted streams. Rather than rely on critical information in corrupted streams to search for headers and perform decoding, contrastive learning directly learns the features of packet headers at different protocol layers and identifies them in a stream sequence. By filtering them out, we can extract the innermost data payload for further analysis. Our evaluation shows that this new approach can successfully recover 71-99\% eavesdropped data hundreds of times faster speed than GSExtract. Besides, the effectiveness of our approach is not largely damaged when stream corruption becomes more severe.
Automatic evaluation of text generation is essential for improving the accuracy of generation tasks. In light of the current trend towards increasingly larger decoder-based language models, we investigate automatic evaluation methods based on such models for text generation. This paper compares various methods, including tuning with encoder-based models and large language models under equal conditions, on two different tasks, machine translation evaluation and semantic textual similarity, in two languages, Japanese and English. Experimental results show that compared to the tuned encoder-based models, the tuned decoder-based models perform poorly. The analysis of the causes for this suggests that the decoder-based models focus on surface word sequences and do not capture meaning. It is also revealed that in-context learning of very large decoder-based models such as ChatGPT makes it difficult to identify fine-grained semantic differences.
Transformers have achieved significant success in medical image segmentation, owing to its capability to capture long-range dependencies. Previous works incorporate convolutional layers into the encoder module of transformers, thereby enhancing their ability to learn local relationships among pixels. However, transformers may suffer from limited generalization capabilities and reduced robustness, attributed to the insufficient spatial recovery ability of their decoders. To address this issue, A convolution sparse vector coding based decoder is proposed , namely CAScaded multi-layer Convolutional Sparse vector Coding DEcoder (CASCSCDE), which represents features extracted by the encoder using sparse vectors. To prove the effectiveness of our CASCSCDE, The widely-used TransUNet model is chosen for the demonstration purpose, and the CASCSCDE is incorporated with TransUNet to establish the TransCASCSCDE architecture. Our experiments demonstrate that TransUNet with CASCSCDE significantly enhances performance on the Synapse benchmark, obtaining up to 3.15\% and 1.16\% improvements in DICE and mIoU scores, respectively. CASCSCDE opens new ways for constructing decoders based on convolutional sparse vector coding.
In recent years, U-Net and its variants have been widely used in pathology image segmentation tasks. One of the key designs of U-Net is the use of skip connections between the encoder and decoder, which helps to recover detailed information after upsampling. While most variations of U-Net adopt the original skip connection design, there is semantic gap between the encoder and decoder that can negatively impact model performance. Therefore, it is important to reduce this semantic gap before conducting skip connection. To address this issue, we propose a new segmentation network called FusionU-Net, which is based on U-Net structure and incorporates a fusion module to exchange information between different skip connections to reduce semantic gaps. Unlike the other fusion modules in existing networks, ours is based on a two-round fusion design that fully considers the local relevance between adjacent encoder layer outputs and the need for bi-directional information exchange across multiple layers. We conducted extensive experiments on multiple pathology image datasets to evaluate our model and found that FusionU-Net achieves better performance compared to other competing methods. We argue our fusion module is more effective than the designs of existing networks, and it could be easily embedded into other networks to further enhance the model performance.
Motivated by recent advances in large language models for Natural Language Processing (NLP), we design a time-series foundation model for forecasting whose out-of-the-box zero-shot performance on a variety of public datasets comes close to the accuracy of state-of-the-art supervised forecasting models for each individual dataset. Our model is based on pretraining a patched-decoder style attention model on a large time-series corpus, and can work well across different forecasting history lengths, prediction lengths and temporal granularities.
Most existing methods for depth estimation from a focal stack of images employ convolutional neural networks (CNNs) using 2D or 3D convolutions over a fixed set of images. However, their effectiveness is constrained by the local properties of CNN kernels, which restricts them to process only focal stacks of fixed number of images during both training and inference. This limitation hampers their ability to generalize to stacks of arbitrary lengths. To overcome these limitations, we present a novel Transformer-based network, FocDepthFormer, which integrates a Transformer with an LSTM module and a CNN decoder. The Transformer's self-attention mechanism allows for the learning of more informative spatial features by implicitly performing non-local cross-referencing. The LSTM module is designed to integrate representations across image stacks of varying lengths. Additionally, we employ multi-scale convolutional kernels in an early-stage encoder to capture low-level features at different degrees of focus/defocus. By incorporating the LSTM, FocDepthFormer can be pre-trained on large-scale monocular RGB depth estimation datasets, improving visual pattern learning and reducing reliance on difficult-to-obtain focal stack data. Extensive experiments on diverse focal stack benchmark datasets demonstrate that our model outperforms state-of-the-art approaches across multiple evaluation metrics.
Self-Supervised Learning (SSL) has shown significant potential and has garnered increasing interest in graph learning. However, particularly for generative SSL methods, its potential in Heterogeneous Graph Learning (HGL) remains relatively underexplored. Generative SSL utilizes an encoder to map the input graph into a latent representation and a decoder to recover the input graph from the latent representation. Previous HGL SSL methods generally design complex strategies to capture graph heterogeneity, which heavily rely on contrastive view construction strategies that are often non-trivial. Yet, refining the latent representation in generative SSL can effectively improve graph learning results. In this study, we propose HGVAE, a generative SSL method specially designed for HGL. Instead of focusing on designing complex strategies to capture heterogeneity, HGVAE centers on refining the latent representation. Specifically, HGVAE innovatively develops a contrastive task based on the latent representation. To ensure the hardness of negative samples, we develop a progressive negative sample generation (PNSG) mechanism that leverages the ability of Variational Inference (VI) to generate high-quality negative samples. As a pioneer in applying generative SSL for HGL, HGVAE refines the latent representation, thereby compelling the model to learn high-quality representations. Compared with various state-of-the-art (SOTA) baselines, HGVAE achieves impressive results, thus validating its superiority.
Coronary artery diseases are among the leading causes of mortality worldwide. Timely and accurate diagnosis, facilitated by precise coronary artery segmentation, is pivotal in changing patient outcomes. In the realm of biomedical imaging, convolutional neural networks, especially the U-Net architecture, have revolutionised segmentation processes. However, one of the primary challenges remains the lack of benchmarking datasets specific to coronary arteries. However through the use of the recently published public dataset ASOCA, the potential of deep learning for accurate coronary segmentation can be improved. This paper delves deep into examining the performance of 25 distinct encoder-decoder combinations. Through analysis of the 40 cases provided to ASOCA participants, it is revealed that the EfficientNet-LinkNet combination, serving as encoder and decoder, stands out. It achieves a Dice coefficient of 0.882 and a 95th percentile Hausdorff distance of 4.753. These findings not only underscore the superiority of our model in comparison to those presented at the MICCAI 2020 challenge but also set the stage for future advancements in coronary artery segmentation, opening doors to enhanced diagnostic and treatment strategies.
Deep learning techniques have implemented many unconditional image generation (UIG) models, such as GAN, Diffusion model, etc. The extremely realistic images (also known as AI-Generated Content, AIGC for short) produced by these models bring urgent needs for intellectual property protection such as data traceability and copyright certification. An attacker can steal the output images of the target model and use them as part of the training data to train a private surrogate UIG model. The implementation mechanisms of UIG models are diverse and complex, and there is no unified and effective protection and verification method at present. To address these issues, we propose a two-stage unified watermark verification mechanism with high-binding effects for such models. In the first stage, we use an encoder to invisibly write the watermark image into the output images of the original AIGC tool, and reversely extract the watermark image through the corresponding decoder. In the second stage, we design the decoder fine-tuning process, and the fine-tuned decoder can make correct judgments on whether the suspicious model steals the original AIGC tool data. Experiments demonstrate our method can complete the verification work with almost zero false positive rate under the condition of only using the model output images. Moreover, the proposed method can achieve data steal verification across different types of UIG models, which further increases the practicality of the method.
Graph neural networks (GNNs) for link prediction can loosely be divided into two broad categories. First, \emph{node-wise} architectures pre-compute individual embeddings for each node that are later combined by a simple decoder to make predictions. While extremely efficient at inference time, model expressiveness is limited such that isomorphic nodes contributing to candidate edges may not be distinguishable, compromising accuracy. In contrast, \emph{edge-wise} methods rely on the formation of edge-specific subgraph embeddings to enrich the representation of pair-wise relationships, disambiguating isomorphic nodes to improve accuracy, but with increased model complexity. To better navigate this trade-off, we propose a novel GNN architecture whereby the \emph{forward pass} explicitly depends on \emph{both} positive (as is typical) and negative (unique to our approach) edges to inform more flexible, yet still cheap node-wise embeddings. This is achieved by recasting the embeddings themselves as minimizers of a forward-pass-specific energy function that favors separation of positive and negative samples. Notably, this energy is distinct from the actual training loss shared by most existing link prediction models, where contrastive pairs only influence the \textit{backward pass}. As demonstrated by extensive empirical evaluations, the resulting architecture retains the inference speed of node-wise models, while producing competitive accuracy with edge-wise alternatives. We released our code at https://github.com/yxzwang/SubmissionverOfYinYanGNN.
Recent studies on semantic communication commonly rely on neural network (NN) based transceivers such as deep joint source and channel coding (DeepJSCC). Unlike traditional transceivers, these neural transceivers are trainable using actual source data and channels, enabling them to extract and communicate semantics. On the flip side, each neural transceiver is inherently biased towards specific source data and channels, making different transceivers difficult to understand intended semantics, particularly upon their initial encounter. To align semantics over multiple neural transceivers, we propose a distributed learning based solution, which leverages split learning (SL) and partial NN fine-tuning techniques. In this method, referred to as SL with layer freezing (SLF), each encoder downloads a misaligned decoder, and locally fine-tunes a fraction of these encoder-decoder NN layers. By adjusting this fraction, SLF controls computing and communication costs. Simulation results confirm the effectiveness of SLF in aligning semantics under different source data and channel dissimilarities, in terms of classification accuracy, reconstruction errors, and recovery time for comprehending intended semantics from misalignment.
Music tagging is a task to predict the tags of music recordings. However, previous music tagging research primarily focuses on close-set music tagging tasks which can not be generalized to new tags. In this work, we propose a zero-shot music tagging system modeled by a joint music and language attention (JMLA) model to address the open-set music tagging problem. The JMLA model consists of an audio encoder modeled by a pretrained masked autoencoder and a decoder modeled by a Falcon7B. We introduce preceiver resampler to convert arbitrary length audio into fixed length embeddings. We introduce dense attention connections between encoder and decoder layers to improve the information flow between the encoder and decoder layers. We collect a large-scale music and description dataset from the internet. We propose to use ChatGPT to convert the raw descriptions into formalized and diverse descriptions to train the JMLA models. Our proposed JMLA system achieves a zero-shot audio tagging accuracy of $ 64.82\% $ on the GTZAN dataset, outperforming previous zero-shot systems and achieves comparable results to previous systems on the FMA and the MagnaTagATune datasets.
We present an end-to-end multichannel speaker-attributed automatic speech recognition (MC-SA-ASR) system that combines a Conformer-based encoder with multi-frame crosschannel attention and a speaker-attributed Transformer-based decoder. To the best of our knowledge, this is the first model that efficiently integrates ASR and speaker identification modules in a multichannel setting. On simulated mixtures of LibriSpeech data, our system reduces the word error rate (WER) by up to 12% and 16% relative compared to previously proposed single-channel and multichannel approaches, respectively. Furthermore, we investigate the impact of different input features, including multichannel magnitude and phase information, on the ASR performance. Finally, our experiments on the AMI corpus confirm the effectiveness of our system for real-world multichannel meeting transcription.
With the emergence of the Segment Anything Model (SAM) as a foundational model for image segmentation, its application has been extensively studied across various domains, including the medical field. However, its potential in the context of histopathology data, specifically in region segmentation, has received relatively limited attention. In this paper, we evaluate SAM's performance in zero-shot and fine-tuned scenarios on histopathology data, with a focus on interactive segmentation. Additionally, we compare SAM with other state-of-the-art interactive models to assess its practical potential and evaluate its generalization capability with domain adaptability. In the experimental results, SAM exhibits a weakness in segmentation performance compared to other models while demonstrating relative strengths in terms of inference time and generalization capability. To improve SAM's limited local refinement ability and to enhance prompt stability while preserving its core strengths, we propose a modification of SAM's decoder. The experimental results suggest that the proposed modification is effective to make SAM useful for interactive histology image segmentation. The code is available at \url{https://github.com/hvcl/SAM_Interactive_Histopathology}
3D object detection models that exploit both LiDAR and camera sensor features are top performers in large-scale autonomous driving benchmarks. A transformer is a popular network architecture used for this task, in which so-called object queries act as candidate objects. Initializing these object queries based on current sensor inputs is a common practice. For this, existing methods strongly rely on LiDAR data however, and do not fully exploit image features. Besides, they introduce significant latency. To overcome these limitations we propose EfficientQ3M, an efficient, modular, and multimodal solution for object query initialization for transformer-based 3D object detection models. The proposed initialization method is combined with a "modality-balanced" transformer decoder where the queries can access all sensor modalities throughout the decoder. In experiments, we outperform the state of the art in transformer-based LiDAR object detection on the competitive nuScenes benchmark and showcase the benefits of input-dependent multimodal query initialization, while being more efficient than the available alternatives for LiDAR-camera initialization. The proposed method can be applied with any combination of sensor modalities as input, demonstrating its modularity.
Structured pruning methods have proven effective in reducing the model size and accelerating inference speed in various network architectures such as Transformers. Despite the versatility of encoder-decoder models in numerous NLP tasks, the structured pruning methods on such models are relatively less explored compared to encoder-only models. In this study, we investigate the behavior of the structured pruning of the encoder-decoder models in the decoupled pruning perspective of the encoder and decoder component, respectively. Our findings highlight two insights: (1) the number of decoder layers is the dominant factor of inference speed, and (2) low sparsity in the pruned encoder network enhances generation quality. Motivated by these findings, we propose a simple and effective framework, NASH, that narrows the encoder and shortens the decoder networks of encoder-decoder models. Extensive experiments on diverse generation and inference tasks validate the effectiveness of our method in both speedup and output quality.
Transformers have become a key architecture in speech processing, but our understanding of how they build up representations of acoustic and linguistic structure is limited. In this study, we address this gap by investigating how measures of 'context-mixing' developed for text models can be adapted and applied to models of spoken language. We identify a linguistic phenomenon that is ideal for such a case study: homophony in French (e.g. livre vs livres), where a speech recognition model has to attend to syntactic cues such as determiners and pronouns in order to disambiguate spoken words with identical pronunciations and transcribe them while respecting grammatical agreement. We perform a series of controlled experiments and probing analyses on Transformer-based speech models. Our findings reveal that representations in encoder-only models effectively incorporate these cues to identify the correct transcription, whereas encoders in encoder-decoder models mainly relegate the task of capturing contextual dependencies to decoder modules.
Transformer-based object detectors (DETR) have shown significant performance across machine vision tasks, ultimately in object detection. This detector is based on a self-attention mechanism along with the transformer encoder-decoder architecture to capture the global context in the image. The critical issue to be addressed is how this model architecture can handle different image nuisances, such as occlusion and adversarial perturbations. We studied this issue by measuring the performance of DETR with different experiments and benchmarking the network with convolutional neural network (CNN) based detectors like YOLO and Faster-RCNN. We found that DETR performs well when it comes to resistance to interference from information loss in occlusion images. Despite that, we found that the adversarial stickers put on the image require the network to produce a new unnecessary set of keys, queries, and values, which in most cases, results in a misdirection of the network. DETR also performed poorer than YOLOv5 in the image corruption benchmark. Furthermore, we found that DETR depends heavily on the main query when making a prediction, which leads to imbalanced contributions between queries since the main query receives most of the gradient flow.
Recent research works have proposed machine learning models for classifying IoT devices connected to a network. However, there is still a practical challenge of not having all devices (and hence their traffic) available during the training of a model. This essentially means, during the operational phase, we need to classify new devices not seen in the training phase. To address this challenge, we propose ZEST -- a ZSL (zero-shot learning) framework based on self-attention for classifying both seen and unseen devices. ZEST consists of i) a self-attention based network feature extractor, termed SANE, for extracting latent space representations of IoT traffic, ii) a generative model that trains a decoder using latent features to generate pseudo data, and iii) a supervised model that is trained on the generated pseudo data for classifying devices. We carry out extensive experiments on real IoT traffic data; our experiments demonstrate i) ZEST achieves significant improvement (in terms of accuracy) over the baselines; ii) SANE is able to better extract meaningful representations than LSTM which has been commonly used for modeling network traffic.
Accurate analysis of microscopy images is hindered by the presence of noise. This noise is usually signal-dependent and often additionally correlated along rows or columns of pixels. Current self- and unsupervised denoisers can address signal-dependent noise, but none can reliably remove noise that is also row- or column-correlated. Here, we present the first fully unsupervised deep learning-based denoiser capable of handling imaging noise that is row-correlated as well as signal-dependent. Our approach uses a Variational Autoencoder (VAE) with a specially designed autoregressive decoder. This decoder is capable of modeling row-correlated and signal-dependent noise but is incapable of independently modeling underlying clean signal. The VAE therefore produces latent variables containing only clean signal information, and these are mapped back into image space using a proposed second decoder network. Our method does not require a pre-trained noise model and can be trained from scratch using unpaired noisy data. We benchmark our approach on microscopy datatsets from a range of imaging modalities and sensor types, each with row- or column-correlated, signal-dependent noise, and show that it outperforms existing self- and unsupervised denoisers.