arXiv subjects

Jian Ren

Publications and source records attributed to Jian Ren.

At least 73 records · Page 4Linked to original sources

arXiv2022

Computer Vision-Aided Reconfigurable Intelligent Surface-Based Beam Tracking: Prototyping and Experimental Results

In this paper, we propose a novel computer vision-based approach to aid Reconfigurable Intelligent Surface (RIS) for dynamic beam tracking and then implement the corresponding prototype verification system. A camera is attached at the RIS to obtain the visual information about the surrounding environment, with which RIS identifies the desired reflected beam direction and then adjusts the reflection coefficients according to the pre-designed codebook. Compared to the conventional approaches that utilize channel estimation or beam sweeping to obtain the reflection coefficients, the proposed one not only saves beam training overhead but also eliminates the requirement for extra feedback links. We build a 20-by-20 RIS running at 5.4 GHz and develop a high-speed control board to ensure the real-time refresh of the reflection coefficients. Meanwhile we implement an independent peer-to-peer communication system to simulate the communication between the base station and the user equipment. The vision-aided RIS prototype system is tested in two mobile scenarios: RIS works in near-field conditions as a passive array antenna of the base station; RIS works in far-field conditions to assist the communication between the base station and the user equipment. The experimental results show that RIS can quickly adjust the reflection coefficients for dynamic beam tracking with the help of visual information.

eess.SP↗

arXiv2022

Are Multimodal Transformers Robust to Missing Modality?

Multimodal data collected from the real world are often imperfect due to missing modalities. Therefore multimodal models that are robust against modal-incomplete data are highly preferred. Recently, Transformer models have shown great success in processing multimodal data. However, existing work has been limited to either architecture designs or pre-training strategies; whether Transformer models are naturally robust against missing-modal data has rarely been investigated. In this paper, we present the first-of-its-kind work to comprehensively investigate the behavior of Transformers in the presence of modal-incomplete data. Unsurprising, we find Transformer models are sensitive to missing modalities while different modal fusion strategies will significantly affect the robustness. What surprised us is that the optimal fusion strategy is dataset dependent even for the same Transformer model; there does not exist a universal strategy that works in general cases. Based on these findings, we propose a principle method to improve the robustness of Transformer models by automatically searching for an optimal fusion strategy regarding input data. Experimental validations on three benchmarks support the superior performance of the proposed method.

cs.CV↗

arXiv2022

Submillimetre galaxies in two massive protoclusters at z = 2.24: witnessing the enrichment of extreme starbursts in the outskirts of HAE density peaks

Submillimetre galaxies represent a rapid growth phase of both star formation and massive galaxies. Mapping SMGs in galaxy protoclusters provides key insights into where and how these extreme starbursts take place in connections with the assembly of the large-scale structure in the early Universe. We search for SMGs at 850$\,μm$ using JCMT/SCUBA-2 in two massive protoclusters at $z=2.24$, BOSS1244 and BOSS1542, and detect 43 and 54 sources with $S_{850}>4\,$mJy at the $4σ$ level within an effective area of 264$\,$arcmin$^2$, respectively. We construct the intrinsic number counts and find that the abundance of SMGs is $2.0\pm0.3$ and $2.1\pm0.2$ times that of the general fields, confirming that BOSS1244 and BOSS1542 contain a higher fraction of dusty galaxies with strongly enhanced star formation. The volume densities of the SMGs are estimated to be $\sim15-$30 times the average, significantly higher than the overdensity factor ($\sim 6$) traced by H$α$ emission-line galaxies (HAEs). More importantly, we discover a prominent offset between the spatial distributions of the two populations in these two protoclusters -- SMGs are mostly located around the high-density regions of HAEs, and few are seen inside these regions. This finding may have revealed for the first time the occurrence of violent star formation enhancement in the outskirts of the HAE density peaks, likely driven by the boosting of gas supplies and/or starburst triggering events. Meanwhile, the lack of SMGs inside the most overdense regions at $z\sim2$ implies a transition to the environment disfavouring extreme starbursts.

astro-ph.GA↗

arXiv2022

Show Me What and Tell Me How: Video Synthesis via Multimodal Conditioning

Most methods for conditional video synthesis use a single modality as the condition. This comes with major limitations. For example, it is problematic for a model conditioned on an image to generate a specific motion trajectory desired by the user since there is no means to provide motion information. Conversely, language information can describe the desired motion, while not precisely defining the content of the video. This work presents a multimodal video generation framework that benefits from text and images provided jointly or separately. We leverage the recent progress in quantized representations for videos and apply a bidirectional transformer with multiple modalities as inputs to predict a discrete video representation. To improve video quality and consistency, we propose a new video token trained with self-learning and an improved mask-prediction algorithm for sampling video tokens. We introduce text augmentation to improve the robustness of the textual representation and diversity of generated videos. Our framework can incorporate various visual modalities, such as segmentation masks, drawings, and partially occluded images. It can generate much longer sequences than the one used for training. In addition, our model can extract visual information as suggested by the text prompt, e.g., "an object in image one is moving northeast", and generate corresponding videos. We run evaluations on three public datasets and a newly collected dataset labeled with facial attributes, achieving state-of-the-art generation results on all four.

cs.CV↗

arXiv2022

F8Net: Fixed-Point 8-bit Only Multiplication for Network Quantization

Neural network quantization is a promising compression technique to reduce memory footprint and save energy consumption, potentially leading to real-time inference. However, there is a performance gap between quantized and full-precision models. To reduce it, existing quantization approaches require high-precision INT32 or full-precision multiplication during inference for scaling or dequantization. This introduces a noticeable cost in terms of memory, speed, and required energy. To tackle these issues, we present F8Net, a novel quantization framework consisting of only fixed-point 8-bit multiplication. To derive our method, we first discuss the advantages of fixed-point multiplication with different formats of fixed-point numbers and study the statistical behavior of the associated fixed-point numbers. Second, based on the statistical and algorithmic analysis, we apply different fixed-point formats for weights and activations of different layers. We introduce a novel algorithm to automatically determine the right format for each layer during training. Third, we analyze a previous quantization algorithm -- parameterized clipping activation (PACT) -- and reformulate it using fixed-point arithmetic. Finally, we unify the recently proposed method for quantization fine-tuning and our fixed-point approach to show the potential of our method. We verify F8Net on ImageNet for MobileNet V1/V2 and ResNet18/50. Our approach achieves comparable and better performance, when compared not only to existing quantization techniques with INT32 multiplication or floating-point arithmetic, but also to the full-precision counterparts, achieving state-of-the-art performance.

cs.CV↗

arXiv2022

A Critical Analysis of Image-based Camera Pose Estimation Techniques

Camera, and associated with its objects within the field of view, localization could benefit many computer vision fields, such as autonomous driving, robot navigation, and augmented reality (AR). In this survey, we first introduce specific application areas and the evaluation metrics for camera localization pose according to different sub-tasks (learning-based 2D-2D task, feature-based 2D-3D task, and 3D-3D task). Then, we review common methods for structure-based camera pose estimation approaches, absolute pose regression and relative pose regression approaches by critically modelling the methods to inspire further improvements in their algorithms such as loss functions, neural network structures. Furthermore, we summarise what are the popular datasets used for camera localization and compare the quantitative and qualitative results of these methods with detailed performance metrics. Finally, we discuss future research possibilities and applications.

cs.CV↗

arXiv2022

Systematic biases in determining dust attenuation curves through galaxy SED fitting

While the slope of the dust attenuation curve ($δ$) is found to correlate with effective dust attenuation ($A_V$) as obtained through spectral energy distribution (SED) fitting, it remains unknown how the fitting degeneracies shape this relation. We examine the degeneracy effects by fitting SEDs of a sample of local star-forming galaxies (SFGs) selected from the Galaxy And Mass Assembly survey, in conjunction with mock galaxy SEDs of known attenuation parameters. A well-designed declining starburst star formation history is adopted to generate model SED templates with intrinsic UV slope ($β_0$) spanning over a reasonably wide range. The best-fitting $β_0$ for our sample SFGs shows a wide coverage, dramatically differing from the limited range of $β_0<-2.2$ for a starburst of constant star formation. Our results show that strong degeneracies between $β_0$, $δ$, and $A_V$ in the SED fitting induce systematic biases leading to a false $A_V$--$δ$ correlation. Our simulation tests reveal that this relationship can be well reproduced even when a flat $A_V$--$δ$ relation is taken to build the input model galaxy SEDs. The variations in best-fitting $δ$ are dominated by the fitting errors. We show that assuming a starburst with constant star formation in SED fitting will result in a steeper attenuation curve, smaller degeneracy errors, and a stronger $A_V$--$δ$ relation. Our findings confirm that the $A_V$--$δ$ relation obtained through SED fitting is likely driven by the systematic biases induced by the fitting degeneracies between $β_0$, $δ$, and $A_V$.

astro-ph.GA↗

arXiv2022

BinarizedAttack: Structural Poisoning Attacks to Graph-based Anomaly Detection

Graph-based Anomaly Detection (GAD) is becoming prevalent due to the powerful representation abilities of graphs as well as recent advances in graph mining techniques. These GAD tools, however, expose a new attacking surface, ironically due to their unique advantage of being able to exploit the relations among data. That is, attackers now can manipulate those relations (i.e., the structure of the graph) to allow some target nodes to evade detection. In this paper, we exploit this vulnerability by designing a new type of targeted structural poisoning attacks to a representative regression-based GAD system termed OddBall. Specially, we formulate the attack against OddBall as a bi-level optimization problem, where the key technical challenge is to efficiently solve the problem in a discrete domain. We propose a novel attack method termed BinarizedAttack based on gradient descent. Comparing to prior arts, BinarizedAttack can better use the gradient information, making it particularly suitable for solving combinatorial optimization problems. Furthermore, we investigate the attack transferability of BinarizedAttack by employing it to attack other representation-learning-based GAD systems. Our comprehensive experiments demonstrate that BinarizedAttack is very effective in enabling target nodes to evade graph-based anomaly detection tools with limited attackers' budget, and in the black-box transfer attack setting, BinarizedAttack is also tested effective and in particular, can significantly change the node embeddings learned by the GAD systems. Our research thus opens the door to studying a new type of attack against security analytic tools that rely on graph data.

cs.LG↗

arXiv2021

The cosmic environment overtakes the local density in shaping galaxy star formation

The gas supply from the cosmic web is the key to sustain star formation in galaxies. It remains to be explored how the cosmic large-scale structure (LSS) effects on galaxy evolution at given local environments. We examine galaxy specific star formation rate as a function of local density in a LSS at $z=0.735$ in the Extended Chandra Deep Field South. The LSS is mapped by 732 galaxies with $R<24$\,mag and redshift at $0.72\le z \le 0.75$ collected from the literature and our spectroscopic observations with Magellan/IMACS, consisting of five galaxy clusters/groups and surrounding filaments over an area of $23.9 \times22.7$\,co-moving\,Mpc$^2$. The spread of spectroscopic redshifts corresponds a velocity dispersion of 494\,km\,s$^{-1}$, indicating the LSS likely to be a thin sheet with a galaxy density $\gtrsim 3.9$ times that of the general field. These clusters/groups in this LSS mostly exhibit elongated morphologies and multiple components connected with surrounding filaments. Strikingly, we find that star-forming galaxies in the LSS keep star formation at the same level as field, and show no dependence on local density but stellar mass. Meanwhile, an increasing fraction of quiescent galaxies is detected at increasing local density in both the LSS and the field, consistent with the expectation that galaxy mass and local dense environment hold the key to quench star formation. Combined together, we conclude that the cosmic environment of the LSS overtakes the local environment in remaining galaxy star formation to the level of the field.

astro-ph.GA↗

arXiv2021

Teachers Do More Than Teach: Compressing Image-to-Image Models

Generative Adversarial Networks (GANs) have achieved huge success in generating high-fidelity images, however, they suffer from low efficiency due to tremendous computational cost and bulky memory usage. Recent efforts on compression GANs show noticeable progress in obtaining smaller generators by sacrificing image quality or involving a time-consuming searching process. In this work, we aim to address these issues by introducing a teacher network that provides a search space in which efficient network architectures can be found, in addition to performing knowledge distillation. First, we revisit the search space of generative models, introducing an inception-based residual block into generators. Second, to achieve target computation cost, we propose a one-step pruning algorithm that searches a student architecture from the teacher model and substantially reduces searching cost. It requires no l1 sparsity regularization and its associated hyper-parameters, simplifying the training procedure. Finally, we propose to distill knowledge through maximizing feature similarity between teacher and student via an index named Global Kernel Alignment (GKA). Our compressed networks achieve similar or even better image fidelity (FID, mIoU) than the original models with much-reduced computational cost, e.g., MACs. Code will be released at https://github.com/snap-research/CAT.

cs.CV↗

arXiv2021

Lottery Ticket Preserves Weight Correlation: Is It Desirable or Not?

In deep model compression, the recent finding "Lottery Ticket Hypothesis" (LTH) (Frankle & Carbin, 2018) pointed out that there could exist a winning ticket (i.e., a properly pruned sub-network together with original weight initialization) that can achieve competitive performance than the original dense network. However, it is not easy to observe such winning property in many scenarios, where for example, a relatively large learning rate is used even if it benefits training the original dense model. In this work, we investigate the underlying condition and rationale behind the winning property, and find that the underlying reason is largely attributed to the correlation between initialized weights and final-trained weights when the learning rate is not sufficiently large. Thus, the existence of winning property is correlated with an insufficient DNN pretraining, and is unlikely to occur for a well-trained DNN. To overcome this limitation, we propose the "pruning & fine-tuning" method that consistently outperforms lottery ticket sparse training under the same pruning algorithm and the same total training epochs. Extensive experiments over multiple deep models (VGG, ResNet, MobileNet-v2) on different datasets have been conducted to justify our proposals.

cs.LG↗

arXiv2021

Flow Guided Transformable Bottleneck Networks for Motion Retargeting

Human motion retargeting aims to transfer the motion of one person in a "driving" video or set of images to another person. Existing efforts leverage a long training video from each target person to train a subject-specific motion transfer model. However, the scalability of such methods is limited, as each model can only generate videos for the given target subject, and such training videos are labor-intensive to acquire and process. Few-shot motion transfer techniques, which only require one or a few images from a target, have recently drawn considerable attention. Methods addressing this task generally use either 2D or explicit 3D representations to transfer motion, and in doing so, sacrifice either accurate geometric modeling or the flexibility of an end-to-end learned representation. Inspired by the Transformable Bottleneck Network, which renders novel views and manipulations of rigid objects, we propose an approach based on an implicit volumetric representation of the image content, which can then be spatially manipulated using volumetric flow fields. We address the challenging question of how to aggregate information across different body poses, learning flow fields that allow for combining content from the appropriate regions of input images of highly non-rigid human subjects performing complex motions into a single implicit volumetric representation. This allows us to learn our 3D representation solely from videos of moving people. Armed with both 3D object understanding and end-to-end learned rendering, this categorically novel representation delivers state-of-the-art image generation quality, as shown by our quantitative and qualitative evaluations.

cs.CV↗

arXiv2021

A Good Image Generator Is What You Need for High-Resolution Video Synthesis

Image and video synthesis are closely related areas aiming at generating content from noise. While rapid progress has been demonstrated in improving image-based models to handle large resolutions, high-quality renderings, and wide variations in image content, achieving comparable video generation results remains problematic. We present a framework that leverages contemporary image generators to render high-resolution videos. We frame the video synthesis problem as discovering a trajectory in the latent space of a pre-trained and fixed image generator. Not only does such a framework render high-resolution videos, but it also is an order of magnitude more computationally efficient. We introduce a motion generator that discovers the desired trajectory, in which content and motion are disentangled. With such a representation, our framework allows for a broad range of applications, including content and motion manipulation. Furthermore, we introduce a new task, which we call cross-domain video synthesis, in which the image and motion generators are trained on disjoint datasets belonging to different domains. This allows for generating moving objects for which the desired video data is not available. Extensive experiments on various datasets demonstrate the advantages of our methods over existing video generation techniques. Code will be released at https://github.com/snap-research/MoCoGAN-HD.

cs.CV↗

arXiv2021

Motion Representations for Articulated Animation

We propose novel motion representations for animating articulated objects consisting of distinct parts. In a completely unsupervised manner, our method identifies object parts, tracks them in a driving video, and infers their motions by considering their principal axes. In contrast to the previous keypoint-based works, our method extracts meaningful and consistent regions, describing locations, shape, and pose. The regions correspond to semantically relevant and distinct object parts, that are more easily detected in frames of the driving video. To force decoupling of foreground from background, we model non-object related global motion with an additional affine transformation. To facilitate animation and prevent the leakage of the shape of the driving object, we disentangle shape and pose of objects in the region space. Our model can animate a variety of objects, surpassing previous methods by a large margin on existing benchmarks. We present a challenging new benchmark with high-resolution videos and show that the improvement is particularly pronounced when articulated objects are considered, reaching 96.6% user preference vs. the state of the art.

cs.CV↗

arXiv2021

In&Out : Diverse Image Outpainting via GAN Inversion

Image outpainting seeks for a semantically consistent extension of the input image beyond its available content. Compared to inpainting -- filling in missing pixels in a way coherent with the neighboring pixels -- outpainting can be achieved in more diverse ways since the problem is less constrained by the surrounding pixels. Existing image outpainting methods pose the problem as a conditional image-to-image translation task, often generating repetitive structures and textures by replicating the content available in the input image. In this work, we formulate the problem from the perspective of inverting generative adversarial networks. Our generator renders micro-patches conditioned on their joint latent code as well as their individual positions in the image. To outpaint an image, we seek for multiple latent codes not only recovering available patches but also synthesizing diverse outpainting by patch-based generation. This leads to richer structure and content in the outpainted regions. Furthermore, our formulation allows for outpainting conditioned on the categorical input, thereby enabling flexible user controls. Extensive experimental results demonstrate the proposed method performs favorably against existing in- and outpainting methods, featuring higher visual quality and diversity.

cs.CV↗

arXiv2021

SMIL: Multimodal Learning with Severely Missing Modality

A common assumption in multimodal learning is the completeness of training data, i.e., full modalities are available in all training examples. Although there exists research endeavor in developing novel methods to tackle the incompleteness of testing data, e.g., modalities are partially missing in testing examples, few of them can handle incomplete training modalities. The problem becomes even more challenging if considering the case of severely missing, e.g., 90% training examples may have incomplete modalities. For the first time in the literature, this paper formally studies multimodal learning with missing modality in terms of flexibility (missing modalities in training, testing, or both) and efficiency (most training data have incomplete modality). Technically, we propose a new method named SMIL that leverages Bayesian meta-learning in uniformly achieving both objectives. To validate our idea, we conduct a series of experiments on three popular benchmarks: MM-IMDb, CMU-MOSI, and avMNIST. The results prove the state-of-the-art performance of SMIL over existing methods and generative baselines including autoencoders and generative adversarial networks. Our code is available at https://github.com/mengmenm/SMIL.

cs.CV↗

arXiv2020

Long Tidal Tails in Merging Galaxies and Their Implications

We investigate the properties of long tidal tails using the largest to date sample of 461 merging galaxies with $\log(M_\ast/\rm M_\odot)\geq9.5$ within $0.2 \leq z \leq 1$ from the COSMOS survey in combination with {\it Hubble Space Telescope} imaging data. Long tidal tails can be briefly divided into three shape types: straight (41\,per\,cent), curved (47\,per\,cent) and plume (12\,per\,cent). Their host galaxies are mostly at late stages of merging, although 31\,per\,cent are galaxy pairs with projected separations $d>20$\,kpc. The high formation rate of straight tidal tails needs to be understood as the projection of curved tidal tails accounts for only a small fraction of the straight tails. We identify 165 tidal dwarf galaxies (TDGs), yielding a TDG production rate of 0.36 per merger. Combined with a galaxy merger fraction and a TDG survival rate from the literature, we estimate that $\sim$\,5\,per\,cent of local dwarf galaxies are of tidal origin, suggesting the tidal formation is not an important formation channel for the dwarf galaxies. More than half of TDGs are located at the tip of their host tails. These TDGs have stellar masses in the range of $7.5\leq\log (M_\ast/\rm M_\odot)\leq9.5$ and appear compact with half-light radii following the $M_\ast$ - $R_{\rm e}$ relation of low-mass elliptical galaxies. However, their surface brightness profiles are generally flatter than those of local disc galaxies. Only 10 out of 165 TDGs have effective radii larger than 1.5 kpc and would qualify as unusually bright ultra-diffuse galaxies.

astro-ph.GA↗

arXiv2020

Identifying stochastic governing equations from data of the most probable transition trajectories

Extracting governing stochastic differential equation models from elusive data is crucial to understand and forecast dynamics for complex systems. We devise a method to extract the drift term and estimate the diffusion coefficient of a governing stochastic dynamical system, from its time-series data of the most probable transition trajectory. By the Onsager-Machlup theory, the most probable transition trajectory satisfies the corresponding Euler-Lagrange equation, which is a second order deterministic ordinary differential equation involving the drift term and diffusion coefficient. We first estimate the coefficients of the Euler-Lagrange equation based on the data of the most probable trajectory, and then we calculate the drift and diffusion coefficients of the governing stochastic dynamical system. These two steps involve sparse regression and optimization. Finally, we illustrate our method with an example and some discussions.

math.NA↗