Searcharxiv⌕ Search

arXiv subjects

Tao Xiang

Publications and source records attributed to Tao Xiang.

At least 163 records · Page 9Linked to original sources
arXiv2023

CLIP for All Things Zero-Shot Sketch-Based Image Retrieval, Fine-Grained or Not

In this paper, we leverage CLIP for zero-shot sketch based image retrieval (ZS-SBIR). We are largely inspired by recent advances on foundation models and the unparalleled generalisation ability they seem to offer, but for the first time tailor it to benefit the sketch community. We put forward novel designs on how best to achieve this synergy, for both the category setting and the fine-grained setting ("all"). At the very core of our solution is a prompt learning setup. First we show just via factoring in sketch-specific prompts, we already have a category-level ZS-SBIR system that overshoots all prior arts, by a large margin (24.8%) - a great testimony on studying the CLIP and ZS-SBIR synergy. Moving onto the fine-grained setup is however trickier, and requires a deeper dive into this synergy. For that, we come up with two specific designs to tackle the fine-grained matching nature of the problem: (i) an additional regularisation loss to ensure the relative separation between sketches and photos is uniform across categories, which is not the case for the gold standard standalone triplet loss, and (ii) a clever patch shuffling technique to help establishing instance-level structural correspondences between sketch-photo pairs. With these designs, we again observe significant performance gains in the region of 26.9% over previous state-of-the-art. The take-home message, if any, is the proposed CLIP and prompt learning paradigm carries great promise in tackling other sketch-related tasks (not limited to ZS-SBIR) where data scarcity remains a great challenge. Project page: https://aneeshan95.github.io/Sketch_LVM/

cs.CV↗
arXiv2023

Multi-Modal Few-Shot Temporal Action Detection

Few-shot (FS) and zero-shot (ZS) learning are two different approaches for scaling temporal action detection (TAD) to new classes. The former adapts a pretrained vision model to a new task represented by as few as a single video per class, whilst the latter requires no training examples by exploiting a semantic description of the new class. In this work, we introduce a new multi-modality few-shot (MMFS) TAD problem, which can be considered as a marriage of FS-TAD and ZS-TAD by leveraging few-shot support videos and new class names jointly. To tackle this problem, we further introduce a novel MUlti-modality PromPt mETa-learning (MUPPET) method. This is enabled by efficiently bridging pretrained vision and language models whilst maximally reusing already learned capacity. Concretely, we construct multi-modal prompts by mapping support videos into the textual token space of a vision-language model using a meta-learned adapter-equipped visual semantics tokenizer. To tackle large intra-class variation, we further design a query feature regulation scheme. Extensive experiments on ActivityNetv1.3 and THUMOS14 demonstrate that our MUPPET outperforms state-of-the-art alternative methods, often by a large margin. We also show that our MUPPET can be easily extended to tackle the few-shot object detection problem and again achieves the state-of-the-art performance on MS-COCO dataset. The code will be available in https://github.com/sauradip/MUPPET

cs.CV↗
arXiv2023

SceneTrilogy: On Human Scene-Sketch and its Complementarity with Photo and Text

In this paper, we extend scene understanding to include that of human sketch. The result is a complete trilogy of scene representation from three diverse and complementary modalities -- sketch, photo, and text. Instead of learning a rigid three-way embedding and be done with it, we focus on learning a flexible joint embedding that fully supports the ``optionality" that this complementarity brings. Our embedding supports optionality on two axes: (i) optionality across modalities -- use any combination of modalities as query for downstream tasks like retrieval, (ii) optionality across tasks -- simultaneously utilising the embedding for either discriminative (e.g., retrieval) or generative tasks (e.g., captioning). This provides flexibility to end-users by exploiting the best of each modality, therefore serving the very purpose behind our proposal of a trilogy in the first place. First, a combination of information-bottleneck and conditional invertible neural networks disentangle the modality-specific component from modality-agnostic in sketch, photo, and text. Second, the modality-agnostic instances from sketch, photo, and text are synergised using a modified cross-attention. Once learned, we show our embedding can accommodate a multi-facet of scene-related tasks, including those enabled for the first time by the inclusion of sketch, all without any task-specific modifications. Project Page: \url{http://www.pinakinathc.me/scenetrilogy}

cs.CV↗
arXiv2023

Exploiting Unlabelled Photos for Stronger Fine-Grained SBIR

This paper advances the fine-grained sketch-based image retrieval (FG-SBIR) literature by putting forward a strong baseline that overshoots prior state-of-the-arts by ~11%. This is not via complicated design though, but by addressing two critical issues facing the community (i) the gold standard triplet loss does not enforce holistic latent space geometry, and (ii) there are never enough sketches to train a high accuracy model. For the former, we propose a simple modification to the standard triplet loss, that explicitly enforces separation amongst photos/sketch instances. For the latter, we put forward a novel knowledge distillation module can leverage photo data for model training. Both modules are then plugged into a novel plug-n-playable training paradigm that allows for more stable training. More specifically, for (i) we employ an intra-modal triplet loss amongst sketches to bring sketches of the same instance closer from others, and one more amongst photos to push away different photo instances while bringing closer a structurally augmented version of the same photo (offering a gain of ~4-6%). To tackle (ii), we first pre-train a teacher on the large set of unlabelled photos over the aforementioned intra-modal photo triplet loss. Then we distill the contextual similarity present amongst the instances in the teacher's embedding space to that in the student's embedding space, by matching the distribution over inter-feature distances of respective samples in both embedding spaces (delivering a further gain of ~4-5%). Apart from outperforming prior arts significantly, our model also yields satisfactory results on generalising to new classes. Project page: https://aneeshan95.github.io/Sketch_PVT/

cs.CV↗
arXiv2023

Generative Model Based Noise Robust Training for Unsupervised Domain Adaptation

Target domain pseudo-labelling has shown effectiveness in unsupervised domain adaptation (UDA). However, pseudo-labels of unlabeled target domain data are inevitably noisy due to the distribution shift between source and target domains. This paper proposes a Generative model-based Noise-Robust Training method (GeNRT), which eliminates domain shift while mitigating label noise. GeNRT incorporates a Distribution-based Class-wise Feature Augmentation (D-CFA) and a Generative-Discriminative classifier Consistency (GDC), both based on the class-wise target distributions modelled by generative models. D-CFA minimizes the domain gap by augmenting the source data with distribution-sampled target features, and trains a noise-robust discriminative classifier by using target domain knowledge from the generative models. GDC regards all the class-wise generative models as generative classifiers and enforces a consistency regularization between the generative and discriminative classifiers. It exploits an ensemble of target knowledge from all the generative models to train a noise-robust discriminative classifier and eventually gets theoretically linked to the Ben-David domain adaptation theorem for reducing the domain gap. Extensive experiments on Office-Home, PACS, and Digit-Five show that our GeNRT achieves comparable performance to state-of-the-art methods under single-source and multi-source UDA settings.

cs.CV↗
arXiv2023

FAME-ViL: Multi-Tasking Vision-Language Model for Heterogeneous Fashion Tasks

In the fashion domain, there exists a variety of vision-and-language (V+L) tasks, including cross-modal retrieval, text-guided image retrieval, multi-modal classification, and image captioning. They differ drastically in each individual input/output format and dataset size. It has been common to design a task-specific model and fine-tune it independently from a pre-trained V+L model (e.g., CLIP). This results in parameter inefficiency and inability to exploit inter-task relatedness. To address such issues, we propose a novel FAshion-focused Multi-task Efficient learning method for Vision-and-Language tasks (FAME-ViL) in this work. Compared with existing approaches, FAME-ViL applies a single model for multiple heterogeneous fashion tasks, therefore being much more parameter-efficient. It is enabled by two novel components: (1) a task-versatile architecture with cross-attention adapters and task-specific adapters integrated into a unified V+L model, and (2) a stable and effective multi-task training strategy that supports learning from heterogeneous data and prevents negative transfer. Extensive experiments on four fashion tasks show that our FAME-ViL can save 61.5% of parameters over alternatives, while significantly outperforming the conventional independently trained single-task models. Code is available at https://github.com/BrandonHanx/FAME-ViL.

cs.CV↗
arXiv2023

The breakdown of both strange metal and superconducting states at a pressure-induced quantum critical point in iron-pnictide superconductors

The strange metal (SM) state, characterized by a linear-in-temperature resistivity, is often seen in the normal state of high temperature superconductors. It is believed that the SM state is one of the keys to understand the underlying mechanism of high-Tc superconductivity. Here we report the first observation of the concurrent breakdown of the SM normal state and superconductivity at a pressure-induced quantum critical point in an iron-pnictide superconductor, Ca10(Pt4As8)((Fe0.97Pt0.03)2As2)5. We find that, upon suppressing the superconducting state by applying pressure, the power exponent changes from 1 to 2, and the corresponding coefficient A, the slope of the temperature-linear resistivity per FeAs layer, gradually diminishes. At a critical pressure (12.5 GPa), A and Tc go to zero concurrently,where a quantum phase transition (QPT) from a superconducting state with a SM normal state to a non-superconducting Fermi liquid state takes place. Scaling analysis on the results obtained from the pressurized 1048 superconductor reveals that A and Tc have a positive relation, which exhibits a similarity with that is seen in other chemically-doped unconventional superconductors, regardless of the type of the tuning method (doping or pressurizing), the crystal structure, the bulk or film superconductors and the nature of dopant. These results suggest that there is a simple but powerful organizational principle of connecting the SM normal state with the high-Tc superconductivity.

cond-mat.supr-con↗
arXiv2023

Post-Processing Temporal Action Detection

Existing Temporal Action Detection (TAD) methods typically take a pre-processing step in converting an input varying-length video into a fixed-length snippet representation sequence, before temporal boundary estimation and action classification. This pre-processing step would temporally downsample the video, reducing the inference resolution and hampering the detection performance in the original temporal resolution. In essence, this is due to a temporal quantization error introduced during the resolution downsampling and recovery. This could negatively impact the TAD performance, but is largely ignored by existing methods. To address this problem, in this work we introduce a novel model-agnostic post-processing method without model redesign and retraining. Specifically, we model the start and end points of action instances with a Gaussian distribution for enabling temporal boundary inference at a sub-snippet level. We further introduce an efficient Taylor-expansion based approximation, dubbed as Gaussian Approximated Post-processing (GAP). Extensive experiments demonstrate that our GAP can consistently improve a wide variety of pre-trained off-the-shelf TAD models on the challenging ActivityNet (+0.2% -0.7% in average mAP) and THUMOS (+0.2% -0.5% in average mAP) benchmarks. Such performance gains are already significant and highly comparable to those achieved by novel model designs. Also, GAP can be integrated with model training for further performance gain. Importantly, GAP enables lower temporal resolutions for more efficient inference, facilitating low-resource applications. The code will be available in https://github.com/sauradip/GAP

cs.CV↗
arXiv2023

Smaller Is Bigger: Rethinking the Embedding Rate of Deep Hiding

Deep hiding, concealing secret information using Deep Neural Networks (DNNs), can significantly increase the embedding rate and improve the efficiency of secret sharing. Existing works mainly force on designing DNNs with higher embedding rates or fancy functionalities. In this paper, we want to answer some fundamental questions: how to increase and what determines the embedding rate of deep hiding. To this end, we first propose a novel Local Deep Hiding (LDH) scheme that significantly increases the embedding rate by hiding large secret images into small local regions of cover images. Our scheme consists of three DNNs: hiding, locating, and revealing. We use the hiding network to convert a secret image in a small imperceptible compact secret code that is embedded into a random local region of a cover image. The locating network assists the revealing process by identifying the position of secret codes in the stego image, while the revealing network recovers all full-size secret images from these identified local regions. Our LDH achieves an extremely high embedding rate, i.e., $16\times24$ bpp and exhibits superior robustness to common image distortions. We also conduct comprehensive experiments to evaluate our scheme under various system settings. We further quantitatively analyze the trade-off between the embedding rate and image quality with different image restoration algorithms.

cs.MM↗
arXiv2023

A Survey of Secure Computation Using Trusted Execution Environments

As an essential technology underpinning trusted computing, the trusted execution environment (TEE) allows one to launch computation tasks on both on- and off-premises data while assuring confidentiality and integrity. This article provides a systematic review and comparison of TEE-based secure computation protocols. We first propose a taxonomy that classifies secure computation protocols into three major categories, namely secure outsourced computation, secure distributed computation and secure multi-party computation. To enable a fair comparison of these protocols, we also present comprehensive assessment criteria with respect to four aspects: setting, methodology, security and performance. Based on these criteria, we review, discuss and compare the state-of-the-art TEE-based secure computation protocols for both general-purpose computation functions and special-purpose ones, such as privacy-preserving machine learning and encrypted database queries. To the best of our knowledge, this article is the first survey to review TEE-based secure computation protocols and the comprehensive comparison can serve as a guideline for selecting suitable protocols for deployment in practice. Finally, we also discuss several future research directions and challenges.

cs.CR↗
arXiv2023

Pressure-induced coevolution of transport properties and lattice stability in CaK(Fe1-xNix)4As4 (x= 0.04 and 0) superconductors with and without spin-vortex crystal state

Here we report the first investigation on correlation between the transport properties and the corresponding stability of the lattice structure for CaK(Fe1-xNix)4As4 (x=0.04 and 0), a new type of putative topological superconductors, with and without a spin-vortex crystal (SVC) state in a wide pressure range involving superconducting to non-superconducting transition and the half- to full-collapse of tetragonal (h-cT and f-cT) phases, by the complementary measurements of high-pressure resistance, Hall coefficient and synchrotron X-ray diffraction. We identify the three critical pressures, P1 that is the turn-on critical pressure of the h-cT phase transition and it coincides with the critical pressure for the sign change of Hall coefficient from positive to negative, a manifestation of the Fermi surface reconstruction, P2 that is the turn-off pressures of the h-cT phase transition, and P3 that is the critical pressure of the f-cT phase transition. By comparing the high-pressure results measured from the two kinds of samples, we find a distinct left-shift of the P1 for the doped sample, at the pressure of which its SVC state is fully suppressed, however the P2 and the P3 remain the same as that of the undoped one. Our results not only provide a consistent understanding on the results reported before, but also demonstrate the importance of the Fe-As bonding in stabilizing the superconductivity of the iron pnictide superconductors through the pressure window.

cond-mat.supr-con↗
arXiv2022

Umklapp scattering in the one-dimensional Hubbard model

The Mott metal-insulator transition is a typical strong correlation effect triggered by the Umklapp scattering. However, in a physical system, the Umklapp scattering coexists with the normal scattering, including both forward and backward scattering, which conserves the total momentum of scattered electrons. Therefore, it is not easy to quantify the contribution of the Umklapp scattering in a Mott metal-insulator transition. To resolve this difficulty, we propose to explore these scattering processes separately. We study the contribution of each scattering process in the one-dimensional Hubbard model using the momentum-space density-matrix renormalization group (kDMRG) and bosonization methods. Our kDMRG calculation confirms that the Mott charge gap results from the Umklapp scattering, but the normal scattering processes strongly renormalize its value. Furthermore, we present a scaling analysis of the Mott charge gap in the bosonization theory and show that the interplay between the Umklapp and forward scattering dictates the charge dynamics in the half-filled Hubbard model.

cond-mat.str-el↗
arXiv2022

Single Stage Multi-Pose Virtual Try-On

Multi-pose virtual try-on (MPVTON) aims to fit a target garment onto a person at a target pose. Compared to traditional virtual try-on (VTON) that fits the garment but keeps the pose unchanged, MPVTON provides a better try-on experience, but is also more challenging due to the dual garment and pose editing objectives. Existing MPVTON methods adopt a pipeline comprising three disjoint modules including a target semantic layout prediction module, a coarse try-on image generator and a refinement try-on image generator. These models are trained separately, leading to sub-optimal model training and unsatisfactory results. In this paper, we propose a novel single stage model for MPVTON. Key to our model is a parallel flow estimation module that predicts the flow fields for both person and garment images conditioned on the target pose. The predicted flows are subsequently used to warp the appearance feature maps of the person and the garment images to construct a style map. The map is then used to modulate the target pose's feature map for target try-on image generation. With the parallel flow estimation design, our model can be trained end-to-end in a single stage and is more computationally efficient, resulting in new SOTA performance on existing MPVTON benchmarks. We further introduce multi-task training and demonstrate that our model can also be applied for traditional VTON and pose transfer tasks and achieve comparable performance to SOTA specialized models on both tasks.

cs.CV↗
arXiv2022

FaD-VLP: Fashion Vision-and-Language Pre-training towards Unified Retrieval and Captioning

Multimodal tasks in the fashion domain have significant potential for e-commerce, but involve challenging vision-and-language learning problems - e.g., retrieving a fashion item given a reference image plus text feedback from a user. Prior works on multimodal fashion tasks have either been limited by the data in individual benchmarks, or have leveraged generic vision-and-language pre-training but have not taken advantage of the characteristics of fashion data. Additionally, these works have mainly been restricted to multimodal understanding tasks. To address these gaps, we make two key contributions. First, we propose a novel fashion-specific pre-training framework based on weakly-supervised triplets constructed from fashion image-text pairs. We show the triplet-based tasks are an effective addition to standard multimodal pre-training tasks. Second, we propose a flexible decoder-based model architecture capable of both fashion retrieval and captioning tasks. Together, our model design and pre-training approach are competitive on a diverse set of fashion tasks, including cross-modal retrieval, image retrieval with text feedback, image captioning, relative image captioning, and multimodal categorization.

cs.CV↗
arXiv2022

Learning to Augment via Implicit Differentiation for Domain Generalization

Machine learning models are intrinsically vulnerable to domain shift between training and testing data, resulting in poor performance in novel domains. Domain generalization (DG) aims to overcome the problem by leveraging multiple source domains to learn a domain-generalizable model. In this paper, we propose a novel augmentation-based DG approach, dubbed AugLearn. Different from existing data augmentation methods, our AugLearn views a data augmentation module as hyper-parameters of a classification model and optimizes the module together with the model via meta-learning. Specifically, at each training step, AugLearn (i) divides source domains into a pseudo source and a pseudo target set, and (ii) trains the augmentation module in such a way that the augmented (synthetic) images can make the model generalize well on the pseudo target set. Moreover, to overcome the expensive second-order gradient computation during meta-learning, we formulate an efficient joint training algorithm, for both the augmentation module and the classification model, based on the implicit function theorem. With the flexibility of augmenting data in both time and frequency spaces, AugLearn shows effectiveness on three standard DG benchmarks, PACS, Office-Home and Digits-DG.

cs.CV↗
arXiv2022

Prediction Calibration for Generalized Few-shot Semantic Segmentation

Generalized Few-shot Semantic Segmentation (GFSS) aims to segment each image pixel into either base classes with abundant training examples or novel classes with only a handful of (e.g., 1-5) training images per class. Compared to the widely studied Few-shot Semantic Segmentation FSS, which is limited to segmenting novel classes only, GFSS is much under-studied despite being more practical. Existing approach to GFSS is based on classifier parameter fusion whereby a newly trained novel class classifier and a pre-trained base class classifier are combined to form a new classifier. As the training data is dominated by base classes, this approach is inevitably biased towards the base classes. In this work, we propose a novel Prediction Calibration Network PCN to address this problem. Instead of fusing the classifier parameters, we fuse the scores produced separately by the base and novel classifiers. To ensure that the fused scores are not biased to either the base or novel classes, a new Transformer-based calibration module is introduced. It is known that the lower-level features are useful of detecting edge information in an input image than higher-level features. Thus, we build a cross-attention module that guides the classifier's final prediction using the fused multi-level features. However, transformers are computationally demanding. Crucially, to make the proposed cross-attention module training tractable at the pixel level, this module is designed based on feature-score cross-covariance and episodically trained to be generalizable at inference time. Extensive experiments on PASCAL-$5^{i}$ and COCO-$20^{i}$ show that our PCN outperforms the state-the-the-art alternatives by large margins.

cs.CV↗
arXiv2022

Electronic Origin of High-Tc Maximization and Persistence in Trilayer Cuprate Superconductors

In high temperature cuprate superconductors, it was found that the superconducting transition temperature Tc depends on the number of CuO2 planes (n) in the structural unit and the maximum Tc is realized in the trilayer system (n=3). It was also found that the trilayer superconductors exhibit an unusual phase diagram that Tc keeps nearly constant in the overdoped region which is in strong contrast to the Tc decrease usually found in other cuprate superconductors. The electronic origin of the Tc maximization in the trilayer superconductors and its high Tc persistence in the overdoped region remains unclear. By taking high resolution laser-based angle resolved photoemission (ARPES) measurements, here we report our revelation of the microscopic origin of the unusual superconducting properties in the trilayer superconductors. For the first time we have observed the trilayer splitting in Bi2Sr2Ca2Cu3O10+d (Bi2223) superconductor. The observed Fermi surface, band structures, superconducting gap and the selective Bogoliubov band hybridizations can be well described by a three-layer interaction model. Quantitative information of the microscopic processes involving intra- and interlayer hoppings and pairings are extracted. The electronic origin of the maximum Tc in Bi2223 and the persistence of the high Tc in the overdoped region is revealed. These results provide key insights in understanding high Tc superconductivity and pave a way to further enhance Tc in the cuprate superconductors.

cond-mat.supr-con↗
arXiv2022

Robust Target Training for Multi-Source Domain Adaptation

Given multiple labeled source domains and a single target domain, most existing multi-source domain adaptation (MSDA) models are trained on data from all domains jointly in one step. Such an one-step approach limits their ability to adapt to the target domain. This is because the training set is dominated by the more numerous and labeled source domain data. The source-domain-bias can potentially be alleviated by introducing a second training step, where the model is fine-tuned with the unlabeled target domain data only using pseudo labels as supervision. However, the pseudo labels are inevitably noisy and when used unchecked can negatively impact the model performance. To address this problem, we propose a novel Bi-level Optimization based Robust Target Training (BORT$^2$) method for MSDA. Given any existing fully-trained one-step MSDA model, BORT$^2$ turns it to a labeling function to generate pseudo-labels for the target data and trains a target model using pseudo-labeled target data only. Crucially, the target model is a stochastic CNN which is designed to be intrinsically robust against label noise generated by the labeling function. Such a stochastic CNN models each target instance feature as a Gaussian distribution with an entropy maximization regularizer deployed to measure the label uncertainty, which is further exploited to alleviate the negative impact of noisy pseudo labels. Training the labeling function and the target model poses a nested bi-level optimization problem, for which we formulate an elegant solution based on implicit differentiation. Extensive experiments demonstrate that our proposed method achieves the state of the art performance on three MSDA benchmarks, including the large-scale DomainNet dataset. Our code will be available at \url{https://github.com/Zhongying-Deng/BORT2}

cs.CV↗