arXiv subjects

Jianlin Su

Publications and source records attributed to Jianlin Su.

34 records · Page 2Linked to original sources

arXiv2021

Sparse-softmax: A Simpler and Faster Alternative Softmax Transformation

The softmax function is widely used in artificial neural networks for the multiclass classification problems, where the softmax transformation enforces the output to be positive and sum to one, and the corresponding loss function allows to use maximum likelihood principle to optimize the model. However, softmax leaves a large margin for loss function to conduct optimizing operation when it comes to high-dimensional classification, which results in low-performance to some extent. In this paper, we provide an empirical study on a simple and concise softmax variant, namely sparse-softmax, to alleviate the problem that occurred in traditional softmax in terms of high-dimensional classification problems. We evaluate our approach in several interdisciplinary tasks, the experimental results show that sparse-softmax is simpler, faster, and produces better results than the baseline models.

cs.LG↗

arXiv2021

BioCopy: A Plug-And-Play Span Copy Mechanism in Seq2Seq Models

Copy mechanisms explicitly obtain unchanged tokens from the source (input) sequence to generate the target (output) sequence under the neural seq2seq framework. However, most of the existing copy mechanisms only consider single word copying from the source sentences, which results in losing essential tokens while copying long spans. In this work, we propose a plug-and-play architecture, namely BioCopy, to alleviate the problem aforementioned. Specifically, in the training stage, we construct a BIO tag for each token and train the original model with BIO tags jointly. In the inference stage, the model will firstly predict the BIO tag at each time step, then conduct different mask strategies based on the predicted BIO label to diminish the scope of the probability distributions over the vocabulary list. Experimental results on two separate generative tasks show that they all outperform the baseline models by adding our BioCopy to the original model structure.

cs.CL↗

arXiv2021

Graph Entropy Guided Node Embedding Dimension Selection for Graph Neural Networks

Graph representation learning has achieved great success in many areas, including e-commerce, chemistry, biology, etc. However, the fundamental problem of choosing the appropriate dimension of node embedding for a given graph still remains unsolved. The commonly used strategies for Node Embedding Dimension Selection (NEDS) based on grid search or empirical knowledge suffer from heavy computation and poor model performance. In this paper, we revisit NEDS from the perspective of minimum entropy principle. Subsequently, we propose a novel Minimum Graph Entropy (MinGE) algorithm for NEDS with graph data. To be specific, MinGE considers both feature entropy and structure entropy on graphs, which are carefully designed according to the characteristics of the rich information in them. The feature entropy, which assumes the embeddings of adjacent nodes to be more similar, connects node features and link topology on graphs. The structure entropy takes the normalized degree as basic unit to further measure the higher-order structure of graphs. Based on them, we design MinGE to directly calculate the ideal node embedding dimension for any graph. Finally, comprehensive experiments with popular Graph Neural Networks (GNNs) on benchmark datasets demonstrate the effectiveness and generalizability of our proposed MinGE.

cs.LG↗

arXiv2021

Whitening Sentence Representations for Better Semantics and Faster Retrieval

Pre-training models such as BERT have achieved great success in many natural language processing tasks. However, how to obtain better sentence representation through these pre-training models is still worthy to exploit. Previous work has shown that the anisotropy problem is an critical bottleneck for BERT-based sentence representation which hinders the model to fully utilize the underlying semantic features. Therefore, some attempts of boosting the isotropy of sentence distribution, such as flow-based model, have been applied to sentence representations and achieved some improvement. In this paper, we find that the whitening operation in traditional machine learning can similarly enhance the isotropy of sentence representations and achieve competitive results. Furthermore, the whitening technique is also capable of reducing the dimensionality of the sentence representation. Our experimental results show that it can not only achieve promising performance but also significantly reduce the storage cost and accelerate the model retrieval speed.

cs.CL↗

arXiv2020

Multi-stage Attention ResU-Net for Semantic Segmentation of Fine-Resolution Remote Sensing Images

The attention mechanism can refine the extracted feature maps and boost the classification performance of the deep network, which has become an essential technique in computer vision and natural language processing. However, the memory and computational costs of the dot-product attention mechanism increase quadratically with the spatio-temporal size of the input. Such growth hinders the usage of attention mechanisms considerably in application scenarios with large-scale inputs. In this Letter, we propose a Linear Attention Mechanism (LAM) to address this issue, which is approximately equivalent to dot-product attention with computational efficiency. Such a design makes the incorporation between attention mechanisms and deep networks much more flexible and versatile. Based on the proposed LAM, we re-factor the skip connections in the raw U-Net and design a Multi-stage Attention ResU-Net (MAResU-Net) for semantic segmentation from fine-resolution remote sensing images. Experiments conducted on the Vaihingen dataset demonstrated the effectiveness and efficiency of our MAResU-Net. Open-source code is available at https://github.com/lironui/Multistage-Attention-ResU-Net.

cs.CV↗

arXiv2020

Multi-Attention-Network for Semantic Segmentation of Fine Resolution Remote Sensing Images

Semantic segmentation of remote sensing images plays an important role in a wide range of applications including land resource management, biosphere monitoring and urban planning. Although the accuracy of semantic segmentation in remote sensing images has been increased significantly by deep convolutional neural networks, several limitations exist in standard models. First, for encoder-decoder architectures such as U-Net, the utilization of multi-scale features causes the underuse of information, where low-level features and high-level features are concatenated directly without any refinement. Second, long-range dependencies of feature maps are insufficiently explored, resulting in sub-optimal feature representations associated with each semantic class. Third, even though the dot-product attention mechanism has been introduced and utilized in semantic segmentation to model long-range dependencies, the large time and space demands of attention impede the actual usage of attention in application scenarios with large-scale input. This paper proposed a Multi-Attention-Network (MANet) to address these issues by extracting contextual dependencies through multiple efficient attention modules. A novel attention mechanism of kernel attention with linear complexity is proposed to alleviate the large computational demand in attention. Based on kernel attention and channel attention, we integrate local feature maps extracted by ResNeXt-101 with their corresponding global dependencies and reweight interdependent channel maps adaptively. Numerical experiments on three large-scale fine resolution remote sensing images captured by different satellite sensors demonstrate the superior performance of the proposed MANet, outperforming the DeepLab V3+, PSPNet, FastFCN, DANet, OCRNet, and other benchmark approaches.

eess.IV↗

arXiv2020

Linear Attention Mechanism: An Efficient Attention for Semantic Segmentation

In this paper, to remedy this deficiency, we propose a Linear Attention Mechanism which is approximate to dot-product attention with much less memory and computational costs. The efficient design makes the incorporation between attention mechanisms and neural networks more flexible and versatile. Experiments conducted on semantic segmentation demonstrated the effectiveness of linear attention mechanism. Code is available at https://github.com/lironui/Linear-Attention-Mechanism.

cs.CV↗

arXiv2020

A Novel Cascade Binary Tagging Framework for Relational Triple Extraction

Extracting relational triples from unstructured text is crucial for large-scale knowledge graph construction. However, few existing works excel in solving the overlapping triple problem where multiple relational triples in the same sentence share the same entities. In this work, we introduce a fresh perspective to revisit the relational triple extraction task and propose a novel cascade binary tagging framework (CasRel) derived from a principled problem formulation. Instead of treating relations as discrete labels as in previous works, our new framework models relations as functions that map subjects to objects in a sentence, which naturally handles the overlapping problem. Experiments show that the CasRel framework already outperforms state-of-the-art methods even when its encoder module uses a randomly initialized BERT encoder, showing the power of the new tagging framework. It enjoys further performance boost when employing a pre-trained BERT encoder, outperforming the strongest baseline by 17.5 and 30.2 absolute gain in F1-score on two public datasets NYT and WebNLG, respectively. In-depth analysis on different scenarios of overlapping triples shows that the method delivers consistent performance gain across all these scenarios. The source code and data are released online.

cs.CL↗

arXiv2020

A Batch Normalized Inference Network Keeps the KL Vanishing Away

Variational Autoencoder (VAE) is widely used as a generative model to approximate a model's posterior on latent variables by combining the amortized variational inference and deep neural networks. However, when paired with strong autoregressive decoders, VAE often converges to a degenerated local optimum known as "posterior collapse". Previous approaches consider the Kullback Leibler divergence (KL) individual for each datapoint. We propose to let the KL follow a distribution across the whole dataset, and analyze that it is sufficient to prevent posterior collapse by keeping the expectation of the KL's distribution positive. Then we propose Batch Normalized-VAE (BN-VAE), a simple but effective approach to set a lower bound of the expectation by regularizing the distribution of the approximate posterior's parameters. Without introducing any new model component or modifying the objective, our approach can avoid the posterior collapse effectively and efficiently. We further show that the proposed BN-VAE can be extended to conditional VAE (CVAE). Empirically, our approach surpasses strong autoregressive baselines on language modeling, text classification and dialogue generation, and rivals more complex approaches while keeping almost the same training time as VAE.

cs.LG↗

arXiv2019

Artist Style Transfer Via Quadratic Potential

In this paper we address the problem of artist style transfer where the painting style of a given artist is applied on a real world photograph. We train our neural networks in adversarial setting via recently introduced quadratic potential divergence for stable learning process. To further improve the quality of generated artist stylized images we also integrate some of the recently introduced deep learning techniques in our method. To our best knowledge this is the first attempt towards artist style transfer via quadratic potential divergence. We provide some stylized image samples in the supplementary material. The source code for experimentation was written in PyTorch and is available online in my GitHub repository.

cs.CV↗

arXiv2019

O-GAN: Extremely Concise Approach for Auto-Encoding Generative Adversarial Networks

In this paper, we propose Orthogonal Generative Adversarial Networks (O-GANs). We decompose the network of discriminator orthogonally and add an extra loss into the objective of common GANs, which can enforce discriminator become an effective encoder. The same extra loss can be embedded into any kind of GANs and there is almost no increase in computation. Furthermore, we discuss the principle of our method, which is relative to the fully-exploiting of the remaining degrees of freedom of discriminator. As we know, our solution is the simplest approach to train a generative adversarial network with auto-encoding ability.

cs.CV↗

arXiv2018

GAN-QP: A Novel GAN Framework without Gradient Vanishing and Lipschitz Constraint

We know SGAN may have a risk of gradient vanishing. A significant improvement is WGAN, with the help of 1-Lipschitz constraint on discriminator to prevent from gradient vanishing. Is there any GAN having no gradient vanishing and no 1-Lipschitz constraint on discriminator? We do find one, called GAN-QP. To construct a new framework of Generative Adversarial Network (GAN) usually includes three steps: 1. choose a probability divergence; 2. convert it into a dual form; 3. play a min-max game. In this articles, we demonstrate that the first step is not necessary. We can analyse the property of divergence and even construct new divergence in dual space directly. As a reward, we obtain a simpler alternative of WGAN: GAN-QP. We demonstrate that GAN-QP have a better performance than WGAN in theory and practice.

cs.LG↗

arXiv2018

Training Generative Adversarial Networks Via Turing Test

In this article, we introduce a new mode for training Generative Adversarial Networks (GANs). Rather than minimizing the distance of evidence distribution $\tilde{p}(x)$ and the generative distribution $q(x)$, we minimize the distance of $\tilde{p}(x_r)q(x_f)$ and $\tilde{p}(x_f)q(x_r)$. This adversarial pattern can be interpreted as a Turing test in GANs. It allows us to use information of real samples during training generator and accelerates the whole training procedure. We even find that just proportionally increasing the size of discriminator and generator, it succeeds on 256x256 resolution without adjusting hyperparameters carefully.

cs.LG↗

arXiv2018

Using deep Residual Networks to search for galaxy-Lyα emitter lens candidates based on spectroscopic-selection

More than one hundred galaxy-scale strong gravitational lens systems have been found by searching for the emission lines coming from galaxies with redshifts higher than the lens galaxies. Based on this spectroscopic-selection method, we introduce the deep Residual Networks (ResNet, a kind of deep Convolutional Neural Networks) to search for the galaxy-Ly$α$ emitter (LAE) lens candidates by recognizing the Ly$α$ emission lines coming from high redshift galaxies ($2 < z < 3$) in the spectra of early-type galaxies (ETGs) at middle redshift ($z\sim 0.5$). The spectra of the ETGs come from the Data Release 12 (DR12) of the Baryon Oscillation Spectroscopic Survey (BOSS) of the Sloan Digital Sky Survey \uppercase\expandafter{\romannumeral3} (SDSS-\uppercase\expandafter{\romannumeral3}). In this paper, we first build a 28 layers ResNet model, and then artificially synthesize 150,000 training spectra, including 140,000 spectra without Ly$α$ lines and 10,000 ones with Ly$α$ lines, to train the networks. After 20 training epochs, we obtain a near-perfect test accuracy at 0.9954. The corresponding loss is 0.0028 and the completeness is 93.6\%. We finally apply our ResNet model to our predictive data with 174 known lens candidates. We obtain 1232 hits including 161 of the 174 known candidates (92.5\% discovery rate). Apart from the hits found in other works, our ResNet model also find 536 new hits. We then perform several subsequent selections on these 536 hits and present 5 most believable lens candidates.

astro-ph.GA↗

arXiv2018

f-VAEs: Improve VAEs with Conditional Flows

In this paper, we integrate VAEs and flow-based generative models successfully and get f-VAEs. Compared with VAEs, f-VAEs generate more vivid images, solved the blurred-image problem of VAEs. Compared with flow-based models such as Glow, f-VAE is more lightweight and converges faster, achieving the same performance under smaller-size architecture.

cs.LG↗

arXiv2018

Variational Inference: A Unified Framework of Generative Models and Some Revelations

We reinterpreting the variational inference in a new perspective. Via this way, we can easily prove that EM algorithm, VAE, GAN, AAE, ALI(BiGAN) are all special cases of variational inference. The proof also reveals the loss of standard GAN is incomplete and it explains why we need to train GAN cautiously. From that, we find out a regularization term to improve stability of GAN training.

cs.LG↗