arXiv subjects

Yi Zhu

Publications and source records attributed to Yi Zhu.

At least 163 records · Page 9Linked to original sources

arXiv2022

NTIRE 2022 Challenge on Efficient Super-Resolution: Methods and Results

This paper reviews the NTIRE 2022 challenge on efficient single image super-resolution with focus on the proposed solutions and results. The task of the challenge was to super-resolve an input image with a magnification factor of $\times$4 based on pairs of low and corresponding high resolution images. The aim was to design a network for single image super-resolution that achieved improvement of efficiency measured according to several metrics including runtime, parameters, FLOPs, activations, and memory consumption while at least maintaining the PSNR of 29.00dB on DIV2K validation set. IMDN is set as the baseline for efficiency measurement. The challenge had 3 tracks including the main track (runtime), sub-track one (model complexity), and sub-track two (overall performance). In the main track, the practical runtime performance of the submissions was evaluated. The rank of the teams were determined directly by the absolute value of the average runtime on the validation set and test set. In sub-track one, the number of parameters and FLOPs were considered. And the individual rankings of the two metrics were summed up to determine a final ranking in this track. In sub-track two, all of the five metrics mentioned in the description of the challenge including runtime, parameter count, FLOPs, activations, and memory consumption were considered. Similar to sub-track one, the rankings of five metrics were summed up to determine a final ranking. The challenge had 303 registered participants, and 43 teams made valid submissions. They gauge the state-of-the-art in efficient single image super-resolution.

cs.CV↗

arXiv2022

Chinese Idiom Paraphrasing

Idioms, are a kind of idiomatic expression in Chinese, most of which consist of four Chinese characters. Due to the properties of non-compositionality and metaphorical meaning, Chinese Idioms are hard to be understood by children and non-native speakers. This study proposes a novel task, denoted as Chinese Idiom Paraphrasing (CIP). CIP aims to rephrase idioms-included sentences to non-idiomatic ones under the premise of preserving the original sentence's meaning. Since the sentences without idioms are easier handled by Chinese NLP systems, CIP can be used to pre-process Chinese datasets, thereby facilitating and improving the performance of Chinese NLP tasks, e.g., machine translation system, Chinese idiom cloze, and Chinese idiom embeddings. In this study, CIP task is treated as a special paraphrase generation task. To circumvent difficulties in acquiring annotations, we first establish a large-scale CIP dataset based on human and machine collaboration, which consists of 115,530 sentence pairs. We further deploy three baselines and two novel CIP approaches to deal with CIP problems. The results show that the proposed methods have better performances than the baselines based on the established CIP dataset.

cs.CL↗

arXiv2022

Gated Multimodal Fusion with Contrastive Learning for Turn-taking Prediction in Human-robot Dialogue

Turn-taking, aiming to decide when the next speaker can start talking, is an essential component in building human-robot spoken dialogue systems. Previous studies indicate that multimodal cues can facilitate this challenging task. However, due to the paucity of public multimodal datasets, current methods are mostly limited to either utilizing unimodal features or simplistic multimodal ensemble models. Besides, the inherent class imbalance in real scenario, e.g. sentence ending with short pause will be mostly regarded as the end of turn, also poses great challenge to the turn-taking decision. In this paper, we first collect a large-scale annotated corpus for turn-taking with over 5,000 real human-robot dialogues in speech and text modalities. Then, a novel gated multimodal fusion mechanism is devised to utilize various information seamlessly for turn-taking prediction. More importantly, to tackle the data imbalance issue, we design a simple yet effective data augmentation method to construct negative instances without supervision and apply contrastive learning to obtain better feature representations. Extensive experiments are conducted and the results demonstrate the superiority and competitiveness of our model over several state-of-the-art baselines.

eess.AS↗

arXiv2022

Large-scale Dynamics of Winds Driven by Line Force from a Thin Accretion Disk

Winds play a significant role in active galactic nuclei feedback process. Previous simulations studying winds only focus on a small dynamical range. Therefore, it is unknown how far the winds can go and what the properties of the winds will be if they can move to large radii. We perform simulations to study the large scale dynamics of winds driven by line force. We find that the properties of the winds depend on both black hole mass ($M_{BH}$) and accretion disk luminosity. When the accretion disk luminosity is $0.6L_{edd}$ ($L_{edd}$ being Eddington luminosity), independent of $M_{BH}$, the winds have kinetic energy flux exceeding $1\% L_{edd}$ and can escape from the black hole potential. For the case with the accretion disk luminosity equaling 0.3$L_{edd}$, the strength of the winds decreases with the decrease of $M_{BH}$. If $M_{BH}$ decreases from $10^9$ to $10^6$ solar mass ($M_\odot$), the winds kinetic energy flux decreases from $\sim 0.01 L_{edd}$ to $ \sim 10^{-6} L_{edd}$. In case of $M_{BH}\geq 10^7 M_\odot$, winds can escape from black hole potential. In the case of $M_{BH}=10^6 M_\odot$, the winds can not escape. We find that for the ultra-fast winds observed in hard X-ray bands (\citealt{Gofford et al. 2015}), the observed dependence of the mass flux and the kinetic energy flux on accretion disk luminosity can be well produced by line force driven winds model. We also find that the properties of the ultra-fast winds observed in soft X-ray bands can be explained by the line force driven winds model.

astro-ph.HE↗

arXiv2022

Prompt-Learning for Short Text Classification

In the short text, the extremely short length, feature sparsity, and high ambiguity pose huge challenges to classification tasks. Recently, as an effective method for tuning Pre-trained Language Models for specific downstream tasks, prompt-learning has attracted a vast amount of attention and research. The main intuition behind the prompt-learning is to insert the template into the input and convert the text classification tasks into equivalent cloze-style tasks. However, most prompt-learning methods expand label words manually or only consider the class name for knowledge incorporating in cloze-style prediction, which will inevitably incur omissions and bias in short text classification tasks. In this paper, we propose a simple short text classification approach that makes use of prompt-learning based on knowledgeable expansion. Taking the special characteristics of short text into consideration, the method can consider both the short text itself and class name during expanding label words space. Specifically, the top $N$ concepts related to the entity in the short text are retrieved from the open Knowledge Graph like Probase, and we further refine the expanded label words by the distance calculation between selected concepts and class labels. Experimental results show that our approach obtains obvious improvement compared with other fine-tuning, prompt-learning, and knowledgeable prompt-tuning methods, outperforming the state-of-the-art by up to 6 Accuracy points on three well-known datasets.

cs.CL↗

arXiv2022

ImpDet: Exploring Implicit Fields for 3D Object Detection

Conventional 3D object detection approaches concentrate on bounding boxes representation learning with several parameters, i.e., localization, dimension, and orientation. Despite its popularity and universality, such a straightforward paradigm is sensitive to slight numerical deviations, especially in localization. By exploiting the property that point clouds are naturally captured on the surface of objects along with accurate location and intensity information, we introduce a new perspective that views bounding box regression as an implicit function. This leads to our proposed framework, termed Implicit Detection or ImpDet, which leverages implicit field learning for 3D object detection. Our ImpDet assigns specific values to points in different local 3D spaces, thereby high-quality boundaries can be generated by classifying points inside or outside the boundary. To solve the problem of sparsity on the object surface, we further present a simple yet efficient virtual sampling strategy to not only fill the empty region, but also learn rich semantic features to help refine the boundaries. Extensive experimental results on KITTI and Waymo benchmarks demonstrate the effectiveness and robustness of unifying implicit fields into object detection.

cs.CV↗

arXiv2022

BigDetection: A Large-scale Benchmark for Improved Object Detector Pre-training

Multiple datasets and open challenges for object detection have been introduced in recent years. To build more general and powerful object detection systems, in this paper, we construct a new large-scale benchmark termed BigDetection. Our goal is to simply leverage the training data from existing datasets (LVIS, OpenImages and Object365) with carefully designed principles, and curate a larger dataset for improved detector pre-training. Specifically, we generate a new taxonomy which unifies the heterogeneous label spaces from different sources. Our BigDetection dataset has 600 object categories and contains over 3.4M training images with 36M bounding boxes. It is much larger in multiple dimensions than previous benchmarks, which offers both opportunities and challenges. Extensive experiments demonstrate its validity as a new benchmark for evaluating different object detection methods, and its effectiveness as a pre-training dataset.

cs.CV↗

arXiv2022

Building Robust Spoken Language Understanding by Cross Attention between Phoneme Sequence and ASR Hypothesis

Building Spoken Language Understanding (SLU) robust to Automatic Speech Recognition (ASR) errors is an essential issue for various voice-enabled virtual assistants. Considering that most ASR errors are caused by phonetic confusion between similar-sounding expressions, intuitively, leveraging the phoneme sequence of speech can complement ASR hypothesis and enhance the robustness of SLU. This paper proposes a novel model with Cross Attention for SLU (denoted as CASLU). The cross attention block is devised to catch the fine-grained interactions between phoneme and word embeddings in order to make the joint representations catch the phonetic and semantic features of input simultaneously and for overcoming the ASR errors in downstream natural language understanding (NLU) tasks. Extensive experiments are conducted on three datasets, showing the effectiveness and competitiveness of our approach. Additionally, We also validate the universality of CASLU and prove its complementarity when combining with other robust SLU techniques.

cs.CL↗

arXiv2022

Global classical solutions of 3D compressible viscoelastic system near equilibrium

In this paper, we prove the global existence of general small solutions to compressible viscoelastic system. We remove the "initial state" assumption ($\tilde ρ_0 \det F_0 =1$) and the "div-curl" structure assumption compared with previous works. It then broadens the class of solutions to a great extent, more precisely the initial density state would not be constant necessarily, and no more structure is need for global well-posedness. It's quite different from the elasticity system in which structure plays an important role. Since we can not obtain any dissipation information for density and deformation tensor, we introduce a new effective flux in the thought of regarding the wildest "nonlinear term" as "linear term". Although the norms of solution may increase now, we can still derive the global existence for it.

math.AP↗

arXiv2021

Contrastive Instruction-Trajectory Learning for Vision-Language Navigation

The vision-language navigation (VLN) task requires an agent to reach a target with the guidance of natural language instruction. Previous works learn to navigate step-by-step following an instruction. However, these works may fail to discriminate the similarities and discrepancies across instruction-trajectory pairs and ignore the temporal continuity of sub-instructions. These problems hinder agents from learning distinctive vision-and-language representations, harming the robustness and generalizability of the navigation policy. In this paper, we propose a Contrastive Instruction-Trajectory Learning (CITL) framework that explores invariance across similar data samples and variance across different ones to learn distinctive representations for robust navigation. Specifically, we propose: (1) a coarse-grained contrastive learning objective to enhance vision-and-language representations by contrasting semantics of full trajectory observations and instructions, respectively; (2) a fine-grained contrastive learning objective to perceive instructions by leveraging the temporal information of the sub-instructions; (3) a pairwise sample-reweighting mechanism for contrastive learning to mine hard samples and hence mitigate the influence of data sampling bias in contrastive learning. Our CITL can be easily integrated with VLN backbones to form a new learning paradigm and achieve better generalizability in unseen environments. Extensive experiments show that the model with CITL surpasses the previous state-of-the-art methods on R2R, R4R, and RxR.

cs.CV↗

arXiv2021

Blending Anti-Aliasing into Vision Transformer

The transformer architectures, based on self-attention mechanism and convolution-free design, recently found superior performance and booming applications in computer vision. However, the discontinuous patch-wise tokenization process implicitly introduces jagged artifacts into attention maps, arising the traditional problem of aliasing for vision transformers. Aliasing effect occurs when discrete patterns are used to produce high frequency or continuous information, resulting in the indistinguishable distortions. Recent researches have found that modern convolution networks still suffer from this phenomenon. In this work, we analyze the uncharted problem of aliasing in vision transformer and explore to incorporate anti-aliasing properties. Specifically, we propose a plug-and-play Aliasing-Reduction Module(ARM) to alleviate the aforementioned issue. We investigate the effectiveness and generalization of the proposed method across multiple tasks and various vision transformer families. This lightweight design consistently attains a clear boost over several famous structures. Furthermore, our module also improves data efficiency and robustness of vision transformers.

cs.CV↗

arXiv2021

SOON: Scenario Oriented Object Navigation with Graph-based Exploration

The ability to navigate like a human towards a language-guided target from anywhere in a 3D embodied environment is one of the 'holy grail' goals of intelligent robots. Most visual navigation benchmarks, however, focus on navigating toward a target from a fixed starting point, guided by an elaborate set of instructions that depicts step-by-step. This approach deviates from real-world problems in which human-only describes what the object and its surrounding look like and asks the robot to start navigation from anywhere. Accordingly, in this paper, we introduce a Scenario Oriented Object Navigation (SOON) task. In this task, an agent is required to navigate from an arbitrary position in a 3D embodied environment to localize a target following a scene description. To give a promising direction to solve this task, we propose a novel graph-based exploration (GBE) method, which models the navigation state as a graph and introduces a novel graph-based exploration approach to learn knowledge from the graph and stabilize training by learning sub-optimal trajectories. We also propose a new large-scale benchmark named From Anywhere to Object (FAO) dataset. To avoid target ambiguity, the descriptions in FAO provide rich semantic scene information includes: object attribute, object relationship, region description, and nearby region description. Our experiments reveal that the proposed GBE outperforms various state-of-the-arts on both FAO and R2R datasets. And the ablation studies on FAO validates the quality of the dataset.

cs.CV↗

arXiv2021

VidTr: Video Transformer Without Convolutions

We introduce Video Transformer (VidTr) with separable-attention for video classification. Comparing with commonly used 3D networks, VidTr is able to aggregate spatio-temporal information via stacked attentions and provide better performance with higher efficiency. We first introduce the vanilla video transformer and show that transformer module is able to perform spatio-temporal modeling from raw pixels, but with heavy memory usage. We then present VidTr which reduces the memory cost by 3.3$\times$ while keeping the same performance. To further optimize the model, we propose the standard deviation based topK pooling for attention ($pool_{topK\_std}$), which reduces the computation by dropping non-informative features along temporal dimension. VidTr achieves state-of-the-art performance on five commonly used datasets with lower computational requirement, showing both the efficiency and effectiveness of our design. Finally, error analysis and visualization show that VidTr is especially good at predicting actions that require long-term temporal reasoning.

cs.CV↗

arXiv2021

Deep Learning for Embodied Vision Navigation: A Survey

"Embodied visual navigation" problem requires an agent to navigate in a 3D environment mainly rely on its first-person observation. This problem has attracted rising attention in recent years due to its wide application in autonomous driving, vacuum cleaner, and rescue robot. A navigation agent is supposed to have various intelligent skills, such as visual perceiving, mapping, planning, exploring and reasoning, etc. Building such an agent that observes, thinks, and acts is a key to real intelligence. The remarkable learning ability of deep learning methods empowered the agents to accomplish embodied visual navigation tasks. Despite this, embodied visual navigation is still in its infancy since a lot of advanced skills are required, including perceiving partially observed visual input, exploring unseen areas, memorizing and modeling seen scenarios, understanding cross-modal instructions, and adapting to a new environment, etc. Recently, embodied visual navigation has attracted rising attention of the community, and numerous works has been proposed to learn these skills. This paper attempts to establish an outline of the current works in the field of embodied visual navigation by providing a comprehensive literature survey. We summarize the benchmarks and metrics, review different methods, analysis the challenges, and highlight the state-of-the-art methods. Finally, we discuss unresolved challenges in the field of embodied visual navigation and give promising directions in pursuing future research.

cs.RO↗

arXiv2021

CrossCLR: Cross-modal Contrastive Learning For Multi-modal Video Representations

Contrastive learning allows us to flexibly define powerful losses by contrasting positive pairs from sets of negative samples. Recently, the principle has also been used to learn cross-modal embeddings for video and text, yet without exploiting its full potential. In particular, previous losses do not take the intra-modality similarities into account, which leads to inefficient embeddings, as the same content is mapped to multiple points in the embedding space. With CrossCLR, we present a contrastive loss that fixes this issue. Moreover, we define sets of highly related samples in terms of their input embeddings and exclude them from the negative samples to avoid issues with false negatives. We show that these principles consistently improve the quality of the learned embeddings. The joint embeddings learned with CrossCLR extend the state of the art in video-text retrieval on Youcook2 and LSMDC datasets and in video captioning on Youcook2 dataset by a large margin. We also demonstrate the generality of the concept by learning improved joint embeddings for other pairs of modalities.

cs.CV↗

arXiv2021

Self-powered InP Nanowire Photodetector for Single Photon Level Detection at Room Temperature

Highly sensitive photodetectors with single photon level detection is one of the key components to a range of emerging technologies, in particular the ever-growing field of optical communication, remote sensing, and quantum computing. Currently, most of the single-photon detection technologies require external biasing at high voltages and/or cooling to low temperatures, posing great limitations for wider applications. Here, we demonstrate InP nanowire array photodetectors that can achieve single-photon level light detection at room temperature without an external bias. We use top-down etched, heavily doped p-type InP nanowires and n-type AZO/ZnO carrier selective contact to form a radial p-n junction with a built-in electric field exceeding 3x10^5 V/cm at 0 V. The device exhibits broadband light sensitivity and can distinguish a single photon per pulse from the dark noise at 0 V, enabled by its design to realize near-ideal broadband absorption, extremely low dark current, and highly efficient charge carrier separation. Meanwhile, the bandwidth of the device reaches above 600 MHz with a timing jitter of 538 ps. The proposed device design provides a new pathway towards low-cost, high-sensitivity, self-powered photodetectors for numerous future applications.

physics.optics↗

arXiv2021

An Unsupervised Method for Building Sentence Simplification Corpora in Multiple Languages

The availability of parallel sentence simplification (SS) is scarce for neural SS modelings. We propose an unsupervised method to build SS corpora from large-scale bilingual translation corpora, alleviating the need for SS supervised corpora. Our method is motivated by the following two findings: neural machine translation model usually tends to generate more high-frequency tokens and the difference of text complexity levels exists between the source and target language of a translation corpus. By taking the pair of the source sentences of translation corpus and the translations of their references in a bridge language, we can construct large-scale pseudo parallel SS data. Then, we keep these sentence pairs with a higher complexity difference as SS sentence pairs. The building SS corpora with an unsupervised approach can satisfy the expectations that the aligned sentences preserve the same meanings and have difference in text complexity levels. Experimental results show that SS methods trained by our corpora achieve the state-of-the-art results and significantly outperform the results on English benchmark WikiLarge.

cs.CL↗

arXiv2021

CrossNorm and SelfNorm for Generalization under Distribution Shifts

Traditional normalization techniques (e.g., Batch Normalization and Instance Normalization) generally and simplistically assume that training and test data follow the same distribution. As distribution shifts are inevitable in real-world applications, well-trained models with previous normalization methods can perform badly in new environments. Can we develop new normalization methods to improve generalization robustness under distribution shifts? In this paper, we answer the question by proposing CrossNorm and SelfNorm. CrossNorm exchanges channel-wise mean and variance between feature maps to enlarge training distribution, while SelfNorm uses attention to recalibrate the statistics to bridge gaps between training and test distributions. CrossNorm and SelfNorm can complement each other, though exploring different directions in statistics usage. Extensive experiments on different fields (vision and language), tasks (classification and segmentation), settings (supervised and semi-supervised), and distribution shift types (synthetic and natural) show the effectiveness. Code is available at https://github.com/amazon-research/crossnorm-selfnorm

cs.CV↗