PyTorchKR Papers

최신 논문

총 9,339편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2304.02838

TBDetector:Transformer-Based Detector for Advanced Persistent Threats with Provenance Graph

APT detection is difficult to detect due to the long-term latency, covert and slow multistage attack patterns of Advanced Persistent Threat (APT). To tackle these issues, we propose TBDetector, a transformer-based advanced persistent threat detection method for APT attack detection. Considering that provenance graphs provide rich historical information and have the powerful attacks historic correlation ability to identify anomalous activities, TBDetector employs provenance analysis for APT detection, which summarizes long-running system execution with space efficiency and utilizes transformer with self-attention based encoder-decoder to extract long-term contextual features of system states to detect slow-acting attacks. Furthermore, we further introduce anomaly scores to investigate the anomaly of different system states, where each state is calculated with an anomaly score corresponding to its similarity score and isolation score. To evaluate the effectiveness of the proposed method, we have conducted experiments on five public datasets, i.e., streamspot, cadets, shellshock, clearscope, and wget_baseline. Experimental results and comparisons with state-of-the-art methods have exhibited better performance of our proposed method.

arXiv:2304.02721

To Asymmetry and Beyond: Structured Pruning of Sequence to Sequence Models for Improved Inference Efficiency

Sequence-to-sequence language models can be used to produce abstractive summaries which are coherent, relevant, and concise. Still, model sizes can make deployment in latency-sensitive or web-scale implementations difficult. This paper studies the relationship between model size, structured pruning, inference efficiency, and summarization accuracy on widely used summarization datasets. We show that model accuracy is tied to the encoder size while inference efficiency is connected to the decoder. Using asymmetric pruning can lead to nearly 3x improvement in inference latency with ~1 point loss in Rouge-2. Moreover, we find both the average degradation and the role of asymmetry to be consistent across model sizes and variations in datasets.

arXiv:2304.02335

Correcting Flaws in Common Disentanglement Metrics

Recent years have seen growing interest in learning disentangled representations, in which distinct features, such as size or shape, are represented by distinct neurons. Quantifying the extent to which a given representation is disentangled is not straightforward; multiple metrics have been proposed. In this paper, we identify two failings of existing metrics, which mean they can assign a high score to a model which is still entangled, and we propose two new metrics, which redress these problems. We then consider the task of compositional generalization. Unlike prior works, we treat this as a classification problem, which allows us to use it to measure the disentanglement ability of the encoder, without depending on the decoder. We show that performance on this task is (a) generally quite poor, (b) correlated with most disentanglement metrics, and (c) most strongly correlated with our newly proposed metrics.

arXiv:2304.02221

Zero-shot domain adaptation of anomalous samples for semi-supervised anomaly detection

Semi-supervised anomaly detection~(SSAD) is a task where normal data and a limited number of anomalous data are available for training. In practical situations, SSAD methods suffer adapting to domain shifts, since anomalous data are unlikely to be available for the target domain in the training phase. To solve this problem, we propose a domain adaptation method for SSAD where no anomalous data are available for the target domain. First, we introduce a domain-adversarial network to a variational auto-encoder-based SSAD model to obtain domain-invariant latent variables. Since the decoder cannot reconstruct the original data solely from domain-invariant latent variables, we conditioned the decoder on the domain label. To compensate for the missing anomalous data of the target domain, we introduce an importance sampling-based weighted loss function that approximates the ideal loss function. Experimental results indicate that the proposed method helps adapt SSAD models to the target domain when no anomalous data are available for the target domain.

arXiv:2304.02211

METransformer: Radiology Report Generation by Transformer with Multiple Learnable Expert Tokens

In clinical scenarios, multi-specialist consultation could significantly benefit the diagnosis, especially for intricate cases. This inspires us to explore a "multi-expert joint diagnosis" mechanism to upgrade the existing "single expert" framework commonly seen in the current literature. To this end, we propose METransformer, a method to realize this idea with a transformer-based backbone. The key design of our method is the introduction of multiple learnable "expert" tokens into both the transformer encoder and decoder. In the encoder, each expert token interacts with both vision tokens and other expert tokens to learn to attend different image regions for image representation. These expert tokens are encouraged to capture complementary information by an orthogonal loss that minimizes their overlap. In the decoder, each attended expert token guides the cross-attention between input words and visual tokens, thus influencing the generated report. A metrics-based expert voting strategy is further developed to generate the final report. By the multi-experts concept, our model enjoys the merits of an ensemble-based approach but through a manner that is computationally more efficient and supports more sophisticated interactions among experts. Experimental results demonstrate the promising performance of our proposed model on two widely used benchmarks. Last but not least, the framework-level innovation makes our work ready to incorporate advances on existing "single-expert" models to further improve its performance.

arXiv:2304.02160

Pac-HuBERT: Self-Supervised Music Source Separation via Primitive Auditory Clustering and Hidden-Unit BERT

In spite of the progress in music source separation research, the small amount of publicly-available clean source data remains a constant limiting factor for performance. Thus, recent advances in self-supervised learning present a largely-unexplored opportunity for improving separation models by leveraging unlabelled music data. In this paper, we propose a self-supervised learning framework for music source separation inspired by the HuBERT speech representation model. We first investigate the potential impact of the original HuBERT model by inserting an adapted version of it into the well-known Demucs V2 time-domain separation model architecture. We then propose a time-frequency-domain self-supervised model, Pac-HuBERT (for primitive auditory clustering HuBERT), that we later use in combination with a Res-U-Net decoder for source separation. Pac-HuBERT uses primitive auditory features of music as unsupervised clustering labels to initialize the self-supervised pretraining process using the Free Music Archive (FMA) dataset. The resulting framework achieves better source-to-distortion ratio (SDR) performance on the MusDB18 test set than the original Demucs V2 and Res-U-Net models. We further demonstrate that it can boost performance with small amounts of supervised data. Ultimately, our proposed framework is an effective solution to the challenge of limited clean source data for music source separation.

arXiv:2304.02633공식 코드 · Papers with Code

HNeRV: A Hybrid Neural Representation for Videos

Implicit neural representations store videos as neural networks and have performed well for various vision tasks such as video compression and denoising. With frame index or positional index as input, implicit representations (NeRV, E-NeRV, \etc) reconstruct video from fixed and content-agnostic embeddings. Such embedding largely limits the regression capacity and internal generalization for video interpolation. In this paper, we propose a Hybrid Neural Representation for Videos (HNeRV), where a learnable encoder generates content-adaptive embeddings, which act as the decoder input. Besides the input embedding, we introduce HNeRV blocks, which ensure model parameters are evenly distributed across the entire network, such that higher layers (layers near the output) can have more capacity to store high-resolution content and video details. With content-adaptive embeddings and re-designed architecture, HNeRV outperforms implicit methods in video regression tasks for both reconstruction quality ($+4.7$ PSNR) and convergence speed ($16\times$ faster), and shows better internal generalization. As a simple and efficient video representation, HNeRV also shows decoding advantages for speed, flexibility, and deployment, compared to traditional codecs~(H.264, H.265) and learning-based compression methods. Finally, we explore the effectiveness of HNeRV on downstream tasks such as video compression and video inpainting. We provide project page at https://haochen-rye.github.io/HNeRV, and Code at https://github.com/haochen-rye/HNeRV

arXiv:2304.01576

MESAHA-Net: Multi-Encoders based Self-Adaptive Hard Attention Network with Maximum Intensity Projections for Lung Nodule Segmentation in CT Scan

Accurate lung nodule segmentation is crucial for early-stage lung cancer diagnosis, as it can substantially enhance patient survival rates. Computed tomography (CT) images are widely employed for early diagnosis in lung nodule analysis. However, the heterogeneity of lung nodules, size diversity, and the complexity of the surrounding environment pose challenges for developing robust nodule segmentation methods. In this study, we propose an efficient end-to-end framework, the multi-encoder-based self-adaptive hard attention network (MESAHA-Net), for precise lung nodule segmentation in CT scans. MESAHA-Net comprises three encoding paths, an attention block, and a decoder block, facilitating the integration of three types of inputs: CT slice patches, forward and backward maximum intensity projection (MIP) images, and region of interest (ROI) masks encompassing the nodule. By employing a novel adaptive hard attention mechanism, MESAHA-Net iteratively performs slice-by-slice 2D segmentation of lung nodules, focusing on the nodule region in each slice to generate 3D volumetric segmentation of lung nodules. The proposed framework has been comprehensively evaluated on the LIDC-IDRI dataset, the largest publicly available dataset for lung nodule segmentation. The results demonstrate that our approach is highly robust for various lung nodule types, outperforming previous state-of-the-art techniques in terms of segmentation accuracy and computational complexity, rendering it suitable for real-time clinical implementation.

arXiv:2304.01611

Q2ATransformer: Improving Medical VQA via an Answer Querying Decoder

Medical Visual Question Answering (VQA) systems play a supporting role to understand clinic-relevant information carried by medical images. The questions to a medical image include two categories: close-end (such as Yes/No question) and open-end. To obtain answers, the majority of the existing medical VQA methods relies on classification approaches, while a few works attempt to use generation approaches or a mixture of the two. The classification approaches are relatively simple but perform poorly on long open-end questions. To bridge this gap, in this paper, we propose a new Transformer based framework for medical VQA (named as Q2ATransformer), which integrates the advantages of both the classification and the generation approaches and provides a unified treatment for the close-end and open-end questions. Specifically, we introduce an additional Transformer decoder with a set of learnable candidate answer embeddings to query the existence of each answer class to a given image-question pair. Through the Transformer attention, the candidate answer embeddings interact with the fused features of the image-question pair to make the decision. In this way, despite being a classification-based approach, our method provides a mechanism to interact with the answer information for prediction like the generation-based approaches. On the other hand, by classification, we mitigate the task difficulty by reducing the search space of answers. Our method achieves new state-of-the-art performance on two medical VQA benchmarks. Especially, for the open-end questions, we achieve 79.19% on VQA-RAD and 54.85% on PathVQA, with 16.09% and 41.45% absolute improvements, respectively.

arXiv:2304.01457공식 코드 · Papers with Code

Exploring Vision-Language Models for Imbalanced Learning

Vision-Language models (VLMs) that use contrastive language-image pre-training have shown promising zero-shot classification performance. However, their performance on imbalanced dataset is relatively poor, where the distribution of classes in the training dataset is skewed, leading to poor performance in predicting minority classes. For instance, CLIP achieved only 5% accuracy on the iNaturalist18 dataset. We propose to add a lightweight decoder to VLMs to avoid OOM (out of memory) problem caused by large number of classes and capture nuanced features for tail classes. Then, we explore improvements of VLMs using prompt tuning, fine-tuning, and incorporating imbalanced algorithms such as Focal Loss, Balanced SoftMax and Distribution Alignment. Experiments demonstrate that the performance of VLMs can be further boosted when used with decoder and imbalanced methods. Specifically, our improved VLMs significantly outperforms zero-shot classification by an average accuracy of 6.58%, 69.82%, and 6.17%, on ImageNet-LT, iNaturalist18, and Places-LT, respectively. We further analyze the influence of pre-training data size, backbones, and training cost. Our study highlights the significance of imbalanced learning algorithms in face of VLMs pre-trained by huge data. We release our code at https://github.com/Imbalance-VLM/Imbalance-VLM.

arXiv:2304.01430

Divided Attention: Unsupervised Multi-Object Discovery with Contextually Separated Slots

We introduce a method to segment the visual field into independently moving regions, trained with no ground truth or supervision. It consists of an adversarial conditional encoder-decoder architecture based on Slot Attention, modified to use the image as context to decode optical flow without attempting to reconstruct the image itself. In the resulting multi-modal representation, one modality (flow) feeds the encoder to produce separate latent codes (slots), whereas the other modality (image) conditions the decoder to generate the first (flow) from the slots. This design frees the representation from having to encode complex nuisance variability in the image due to, for instance, illumination and reflectance properties of the scene. Since customary autoencoding based on minimizing the reconstruction error does not preclude the entire flow from being encoded into a single slot, we modify the loss to an adversarial criterion based on Contextual Information Separation. The resulting min-max optimization fosters the separation of objects and their assignment to different attention slots, leading to Divided Attention, or DivA. DivA outperforms recent unsupervised multi-object motion segmentation methods while tripling run-time speed up to 104FPS and reducing the performance gap from supervised methods to 12% or less. DivA can handle different numbers of objects and different image sizes at training and test time, is invariant to permutation of object labels, and does not require explicit regularization.

arXiv:2304.01106

Crossword: A Semantic Approach to Data Compression via Masking

The traditional methods for data compression are typically based on the symbol-level statistics, with the information source modeled as a long sequence of i.i.d. random variables or a stochastic process, thus establishing the fundamental limit as entropy for lossless compression and as mutual information for lossy compression. However, the source (including text, music, and speech) in the real world is often statistically ill-defined because of its close connection to human perception, and thus the model-driven approach can be quite suboptimal. This study places careful emphasis on English text and exploits its semantic aspect to enhance the compression efficiency further. The main idea stems from the puzzle crossword, observing that the hidden words can still be precisely reconstructed so long as some key letters are provided. The proposed masking-based strategy resembles the above game. In a nutshell, the encoder evaluates the semantic importance of each word according to the semantic loss and then masks the minor ones, while the decoder aims to recover the masked words from the semantic context by means of the Transformer. Our experiments show that the proposed semantic approach can achieve much higher compression efficiency than the traditional methods such as Huffman code and UTF-8 code, while preserving the meaning in the target text to a great extent.

arXiv:2304.00171

Practical Conformer: Optimizing size, speed and flops of Conformer for on-Device and cloud ASR

Conformer models maintain a large number of internal states, the vast majority of which are associated with self-attention layers. With limited memory bandwidth, reading these from memory at each inference step can slow down inference. In this paper, we design an optimized conformer that is small enough to meet on-device restrictions and has fast inference on TPUs. We explore various ideas to improve the execution speed, including replacing lower conformer blocks with convolution-only blocks, strategically downsizing the architecture, and utilizing an RNNAttention-Performer. Our optimized conformer can be readily incorporated into a cascaded-encoder setting, allowing a second-pass decoder to operate on its output and improve the accuracy whenever more resources are available. Altogether, we find that these optimizations can reduce latency by a factor of 6.8x, and come at a reasonable trade-off in quality. With the cascaded second-pass, we show that the recognition accuracy is completely recoverable. Thus, our proposed encoder can double as a strong standalone encoder in on device, and as the first part of a high-performance ASR pipeline.

arXiv:2304.00335

Volumetric Attribute Compression for 3D Point Clouds using Feedforward Network with Geometric Attention

We study 3D point cloud attribute compression using a volumetric approach: given a target volumetric attribute function $f : \mathbb{R}^3 \rightarrow \mathbb{R}$, we quantize and encode parameter vector $\theta$ that characterizes $f$ at the encoder, for reconstruction $f_{\hat{\theta}}(\mathbf{x})$ at known 3D points $\mathbf{x}$'s at the decoder. Extending a previous work Region Adaptive Hierarchical Transform (RAHT) that employs piecewise constant functions to span a nested sequence of function spaces, we propose a feedforward linear network that implements higher-order B-spline bases spanning function spaces without eigen-decomposition. Feedforward network architecture means that the system is amenable to end-to-end neural learning. The key to our network is space-varying convolution, similar to a graph operator, whose weights are computed from the known 3D geometry for normalization. We show that the number of layers in the normalization at the encoder is equivalent to the number of terms in a matrix inverse Taylor series. Experimental results on real-world 3D point clouds show up to 2-3 dB gain over RAHT in energy compaction and 20-30% bitrate reduction.

arXiv:2304.01184공식 코드 · Papers with Code

WeakTr: Exploring Plain Vision Transformer for Weakly-supervised Semantic Segmentation

This paper explores the properties of the plain Vision Transformer (ViT) for Weakly-supervised Semantic Segmentation (WSSS). The class activation map (CAM) is of critical importance for understanding a classification network and launching WSSS. We observe that different attention heads of ViT focus on different image areas. Thus a novel weight-based method is proposed to end-to-end estimate the importance of attention heads, while the self-attention maps are adaptively fused for high-quality CAM results that tend to have more complete objects. Besides, we propose a ViT-based gradient clipping decoder for online retraining with the CAM results to complete the WSSS task. We name this plain Transformer-based Weakly-supervised learning framework WeakTr. It achieves the state-of-the-art WSSS performance on standard benchmarks, i.e., 78.4% mIoU on the val set of PASCAL VOC 2012 and 50.3% mIoU on the val set of COCO 2014. Code is available at https://github.com/hustvl/WeakTr.

arXiv:2304.01171공식 코드 · Papers with Code

Revisiting Context Aggregation for Image Matting

Traditional studies emphasize the significance of context information in improving matting performance. Consequently, deep learning-based matting methods delve into designing pooling or affinity-based context aggregation modules to achieve superior results. However, these modules cannot well handle the context scale shift caused by the difference in image size during training and inference, resulting in matting performance degradation. In this paper, we revisit the context aggregation mechanisms of matting networks and find that a basic encoder-decoder network without any context aggregation modules can actually learn more universal context aggregation, thereby achieving higher matting performance compared to existing methods. Building on this insight, we present AEMatter, a matting network that is straightforward yet very effective. AEMatter adopts a Hybrid-Transformer backbone with appearance-enhanced axis-wise learning (AEAL) blocks to build a basic network with strong context aggregation learning capability. Furthermore, AEMatter leverages a large image training strategy to assist the network in learning context aggregation from data. Extensive experiments on five popular matting datasets demonstrate that the proposed AEMatter outperforms state-of-the-art matting methods by a large margin.

arXiv:2304.01064공식 코드 · Papers with Code

Real-time 6K Image Rescaling with Rate-distortion Optimization

Contemporary image rescaling aims at embedding a high-resolution (HR) image into a low-resolution (LR) thumbnail image that contains embedded information for HR image reconstruction. Unlike traditional image super-resolution, this enables high-fidelity HR image restoration faithful to the original one, given the embedded information in the LR thumbnail. However, state-of-the-art image rescaling methods do not optimize the LR image file size for efficient sharing and fall short of real-time performance for ultra-high-resolution (e.g., 6K) image reconstruction. To address these two challenges, we propose a novel framework (HyperThumbnail) for real-time 6K rate-distortion-aware image rescaling. Our framework first embeds an HR image into a JPEG LR thumbnail by an encoder with our proposed quantization prediction module, which minimizes the file size of the embedding LR JPEG thumbnail while maximizing HR reconstruction quality. Then, an efficient frequency-aware decoder reconstructs a high-fidelity HR image from the LR one in real time. Extensive experiments demonstrate that our framework outperforms previous image rescaling baselines in rate-distortion performance and can perform 6K image reconstruction in real time.

arXiv:2304.00495

Multimodal Hyperspectral Image Classification via Interconnected Fusion

Existing multiple modality fusion methods, such as concatenation, summation, and encoder-decoder-based fusion, have recently been employed to combine modality characteristics of Hyperspectral Image (HSI) and Light Detection And Ranging (LiDAR). However, these methods consider the relationship of HSI-LiDAR signals from limited perspectives. More specifically, they overlook the contextual information across modalities of HSI and LiDAR and the intra-modality characteristics of LiDAR. In this paper, we provide a new insight into feature fusion to explore the relationships across HSI and LiDAR modalities comprehensively. An Interconnected Fusion (IF) framework is proposed. Firstly, the center patch of the HSI input is extracted and replicated to the size of the HSI input. Then, nine different perspectives in the fusion matrix are generated by calculating self-attention and cross-attention among the replicated center patch, HSI input, and corresponding LiDAR input. In this way, the intra- and inter-modality characteristics can be fully exploited, and contextual information is considered in both intra-modality and inter-modality manner. These nine interrelated elements in the fusion matrix can complement each other and eliminate biases, which can generate a multi-modality representation for classification accurately. Extensive experiments have been conducted on three widely used datasets: Trento, MUUFL, and Houston. The IF framework achieves state-of-the-art results on these datasets compared to existing approaches.

arXiv:2304.00173

Lego-Features: Exporting modular encoder features for streaming and deliberation ASR

In end-to-end (E2E) speech recognition models, a representational tight-coupling inevitably emerges between the encoder and the decoder. We build upon recent work that has begun to explore building encoders with modular encoded representations, such that encoders and decoders from different models can be stitched together in a zero-shot manner without further fine-tuning. While previous research only addresses full-context speech models, we explore the problem in a streaming setting as well. Our framework builds on top of existing encoded representations, converting them to modular features, dubbed as Lego-Features, without modifying the pre-trained model. The features remain interchangeable when the model is retrained with distinct initializations. Though sparse, we show that the Lego-Features are powerful when tested with RNN-T or LAS decoders, maintaining high-quality downstream performance. They are also rich enough to represent the first-pass prediction during two-pass deliberation. In this scenario, they outperform the N-best hypotheses, since they do not need to be supplemented with acoustic features to deliver the best results. Moreover, generating the Lego-Features does not require beam search or auto-regressive computation. Overall, they present a modular, powerful and cheap alternative to the standard encoder output, as well as the N-best hypotheses.

arXiv:2303.18219

SemHint-MD: Learning from Noisy Semantic Labels for Self-Supervised Monocular Depth Estimation

Without ground truth supervision, self-supervised depth estimation can be trapped in a local minimum due to the gradient-locality issue of the photometric loss. In this paper, we present a framework to enhance depth by leveraging semantic segmentation to guide the network to jump out of the local minimum. Prior works have proposed to share encoders between these two tasks or explicitly align them based on priors like the consistency between edges in the depth and segmentation maps. Yet, these methods usually require ground truth or high-quality pseudo labels, which may not be easily accessible in real-world applications. In contrast, we investigate self-supervised depth estimation along with a segmentation branch that is supervised with noisy labels provided by models pre-trained with limited data. We extend parameter sharing from the encoder to the decoder and study the influence of different numbers of shared decoder parameters on model performance. Also, we propose to use cross-task information to refine current depth and segmentation predictions to generate pseudo-depth and semantic labels for training. The advantages of the proposed method are demonstrated through extensive experiments on the KITTI benchmark and a downstream task for endoscopic tissue deformation tracking.