Searcharxiv⌕ Search

arXiv subjects

Heng Wang

Publications and source records attributed to Heng Wang.

At least 145 records · Page 8Linked to original sources
arXiv2024

Finetuned Multimodal Language Models Are High-Quality Image-Text Data Filters

We propose a novel framework for filtering image-text data by leveraging fine-tuned Multimodal Language Models (MLMs). Our approach outperforms predominant filtering methods (e.g., CLIPScore) via integrating the recent advances in MLMs. We design four distinct yet complementary metrics to holistically measure the quality of image-text data. A new pipeline is established to construct high-quality instruction data for fine-tuning MLMs as data filters. Comparing with CLIPScore, our MLM filters produce more precise and comprehensive scores that directly improve the quality of filtered data and boost the performance of pre-trained models. We achieve significant improvements over CLIPScore on popular foundation models (i.e., CLIP and BLIP2) and various downstream tasks. Our MLM filter can generalize to different models and tasks, and be used as a drop-in replacement for CLIPScore. An additional ablation study is provided to verify our design choices for the MLM filter.

cs.CV↗
arXiv2024

Can Language Models Solve Graph Problems in Natural Language?

Large language models (LLMs) are increasingly adopted for a variety of tasks with implicit graphical structures, such as planning in robotics, multi-hop question answering or knowledge probing, structured commonsense reasoning, and more. While LLMs have advanced the state-of-the-art on these tasks with structure implications, whether LLMs could explicitly process textual descriptions of graphs and structures, map them to grounded conceptual spaces, and perform structured operations remains underexplored. To this end, we propose NLGraph (Natural Language Graph), a comprehensive benchmark of graph-based problem solving designed in natural language. NLGraph contains 29,370 problems, covering eight graph reasoning tasks with varying complexity from simple tasks such as connectivity and shortest path up to complex problems such as maximum flow and simulating graph neural networks. We evaluate LLMs (GPT-3/4) with various prompting approaches on the NLGraph benchmark and find that 1) language models do demonstrate preliminary graph reasoning abilities, 2) the benefit of advanced prompting and in-context learning diminishes on more complex graph problems, while 3) LLMs are also (un)surprisingly brittle in the face of spurious correlations in graph and problem settings. We then propose Build-a-Graph Prompting and Algorithmic Prompting, two instruction-based approaches to enhance LLMs in solving natural language graph problems. Build-a-Graph and Algorithmic prompting improve the performance of LLMs on NLGraph by 3.07% to 16.85% across multiple tasks and settings, while how to solve the most complicated graph reasoning tasks in our setup with language models remains an open research question. The NLGraph benchmark and evaluation code are available at https://github.com/Arthur-Heng/NLGraph.

cs.CL↗
arXiv2023

Video Recognition in Portrait Mode

The creation of new datasets often presents new challenges for video recognition and can inspire novel ideas while addressing these challenges. While existing datasets mainly comprise landscape mode videos, our paper seeks to introduce portrait mode videos to the research community and highlight the unique challenges associated with this video format. With the growing popularity of smartphones and social media applications, recognizing portrait mode videos is becoming increasingly important. To this end, we have developed the first dataset dedicated to portrait mode video recognition, namely PortraitMode-400. The taxonomy of PortraitMode-400 was constructed in a data-driven manner, comprising 400 fine-grained categories, and rigorous quality assurance was implemented to ensure the accuracy of human annotations. In addition to the new dataset, we conducted a comprehensive analysis of the impact of video format (portrait mode versus landscape mode) on recognition accuracy and spatial bias due to the different formats. Furthermore, we designed extensive experiments to explore key aspects of portrait mode video recognition, including the choice of data augmentation, evaluation procedure, the importance of temporal information, and the role of audio modality. Building on the insights from our experimental results and the introduction of PortraitMode-400, our paper aims to inspire further research efforts in this emerging research area.

cs.CV↗
arXiv2023

V2A-Mapper: A Lightweight Solution for Vision-to-Audio Generation by Connecting Foundation Models

Building artificial intelligence (AI) systems on top of a set of foundation models (FMs) is becoming a new paradigm in AI research. Their representative and generative abilities learnt from vast amounts of data can be easily adapted and transferred to a wide range of downstream tasks without extra training from scratch. However, leveraging FMs in cross-modal generation remains under-researched when audio modality is involved. On the other hand, automatically generating semantically-relevant sound from visual input is an important problem in cross-modal generation studies. To solve this vision-to-audio (V2A) generation problem, existing methods tend to design and build complex systems from scratch using modestly sized datasets. In this paper, we propose a lightweight solution to this problem by leveraging foundation models, specifically CLIP, CLAP, and AudioLDM. We first investigate the domain gap between the latent space of the visual CLIP and the auditory CLAP models. Then we propose a simple yet effective mapper mechanism (V2A-Mapper) to bridge the domain gap by translating the visual input between CLIP and CLAP spaces. Conditioned on the translated CLAP embedding, pretrained audio generative FM AudioLDM is adopted to produce high-fidelity and visually-aligned sound. Compared to previous approaches, our method only requires a quick training of the V2A-Mapper. We further analyze and conduct extensive experiments on the choice of the V2A-Mapper and show that a generative mapper is better at fidelity and variability (FD) while a regression mapper is slightly better at relevance (CS). Both objective and subjective evaluation on two V2A datasets demonstrate the superiority of our proposed method compared to current state-of-the-art approaches - trained with 86% fewer parameters but achieving 53% and 19% improvement in FD and CS, respectively.

cs.CV↗
arXiv2023

InfiMM-Eval: Complex Open-Ended Reasoning Evaluation For Multi-Modal Large Language Models

Multi-modal Large Language Models (MLLMs) are increasingly prominent in the field of artificial intelligence. These models not only excel in traditional vision-language tasks but also demonstrate impressive performance in contemporary multi-modal benchmarks. Although many of these benchmarks attempt to holistically evaluate MLLMs, they typically concentrate on basic reasoning tasks, often yielding only simple yes/no or multi-choice responses. These methods naturally lead to confusion and difficulties in conclusively determining the reasoning capabilities of MLLMs. To mitigate this issue, we manually curate a benchmark dataset specifically designed for MLLMs, with a focus on complex reasoning tasks. Our benchmark comprises three key reasoning categories: deductive, abductive, and analogical reasoning. The queries in our dataset are intentionally constructed to engage the reasoning capabilities of MLLMs in the process of generating answers. For a fair comparison across various MLLMs, we incorporate intermediate reasoning steps into our evaluation criteria. In instances where an MLLM is unable to produce a definitive answer, its reasoning ability is evaluated by requesting intermediate reasoning steps. If these steps align with our manual annotations, appropriate scores are assigned. This evaluation scheme resembles methods commonly used in human assessments, such as exams or assignments, and represents what we consider a more effective assessment technique compared with existing benchmarks. We evaluate a selection of representative MLLMs using this rigorously developed open-ended multi-step elaborate reasoning benchmark, designed to challenge and accurately measure their reasoning capabilities. The code and data will be released at https://infimm.github.io/InfiMM-Eval/

cs.CV↗
arXiv2023

An Ultra-fast Quantum Random Number Generation Scheme Based on Laser Phase Noise

Based on the intrinsic random property of quantum mechanics, quantum random number generators allow for access of truly unpredictable random sequence and are now heading towards high performance and small miniaturization, among which a popular scheme is based on the laser phase noise. However, this scheme is generally limited in speed and implementation complexity, especially for chip integration. In this work, a general physical model based on wiener process for such schemes is introduced, which provides an approach to clearly explain the limitation on the generation rate and comprehensively optimize the system performance. We present an insight to exploit the potential bandwidth of the quantum entropy source that contains plentiful quantum randomness with a simple spectral filtering method and experimentally boost the bandwidth of the corresponding quantum entropy source to 20 GHz, based on which an ultra-fast generation rate of 218 Gbps is demonstrated, setting a new record for laser phase noise based schemes by one order of magnitude. Our proposal significantly enhances the ceiling speed of such schemes without requiring extra complex hardware, thus effectively benefits the corresponding chip integration with high performance and low implementation cost, which paves the way for its large-scale applications.

quant-ph↗
arXiv2023

Prominent Josephson tunneling between twisted single copper oxide planes of Bi$_2$Sr$_{2-x}$LaxCuO$_{6+y}$

Josephson tunneling in twisted cuprate junctions provides a litmus test for the pairing symmetry, which is fundamental for understanding the microscopic mechanism of high temperature superconductivity. This issue is rekindled by experimental advances in van der Waals stacking and the proposal of an emergent d+id-wave. So far, all experiments have been carried out on Bi$_2$Sr$_2$CaCu$_2$O$_{8+x}$ (Bi-2212) with double CuO$_2$ planes but show controversial results. Here, we investigate junctions made of Bi$_2$Sr$_{2-x}$La$_x$CuO$_{6+y}$ (Bi-2201) with single CuO$_2$ planes. Our on-site cold stacking technique ensures uncompromised crystalline quality and stoichiometry at the interface. Junctions with carefully calibrated twist angles around 45° show strong Josephson tunneling and conventional temperature dependence. Furthermore, we observe standard Fraunhofer diffraction patterns and integer Fiske steps in a junction with a twist angle of 45.0$\pm$0.2°. Together, these results pose strong constraints on the d or d+id-wave pairing and suggest an indispensable isotropic pairing component.

cond-mat.supr-con↗
arXiv2023

GPT-4V(ision) as a Generalist Evaluator for Vision-Language Tasks

Automatically evaluating vision-language tasks is challenging, especially when it comes to reflecting human judgments due to limitations in accounting for fine-grained details. Although GPT-4V has shown promising results in various multi-modal tasks, leveraging GPT-4V as a generalist evaluator for these tasks has not yet been systematically explored. We comprehensively validate GPT-4V's capabilities for evaluation purposes, addressing tasks ranging from foundational image-to-text and text-to-image synthesis to high-level image-to-image translations and multi-images to text alignment. We employ two evaluation methods, single-answer grading and pairwise comparison, using GPT-4V. Notably, GPT-4V shows promising agreement with humans across various tasks and evaluation methods, demonstrating immense potential for multi-modal LLMs as evaluators. Despite limitations like restricted visual clarity grading and real-world complex reasoning, its ability to provide human-aligned scores enriched with detailed explanations is promising for universal automatic evaluator.

cs.CV↗
arXiv2023

Detecting Spoilers in Movie Reviews with External Movie Knowledge and User Networks

Online movie review platforms are providing crowdsourced feedback for the film industry and the general public, while spoiler reviews greatly compromise user experience. Although preliminary research efforts were made to automatically identify spoilers, they merely focus on the review content itself, while robust spoiler detection requires putting the review into the context of facts and knowledge regarding movies, user behavior on film review platforms, and more. In light of these challenges, we first curate a large-scale network-based spoiler detection dataset LCS and a comprehensive and up-to-date movie knowledge base UKM. We then propose MVSD, a novel Multi-View Spoiler Detection framework that takes into account the external knowledge about movies and user activities on movie review platforms. Specifically, MVSD constructs three interconnecting heterogeneous information networks to model diverse data sources and their multi-view attributes, while we design and employ a novel heterogeneous graph neural network architecture for spoiler detection as node-level classification. Extensive experiments demonstrate that MVSD advances the state-of-the-art on two spoiler detection datasets, while the introduction of external knowledge and user interactions help ground robust spoiler detection. Our data and code are available at https://github.com/Arthur-Heng/Spoiler-Detection

cs.AI↗
arXiv2023

The Devil is in the Details: A Deep Dive into the Rabbit Hole of Data Filtering

The quality of pre-training data plays a critical role in the performance of foundation models. Popular foundation models often design their own recipe for data filtering, which makes it hard to analyze and compare different data filtering approaches. DataComp is a new benchmark dedicated to evaluating different methods for data filtering. This paper describes our learning and solution when participating in the DataComp challenge. Our filtering strategy includes three stages: single-modality filtering, cross-modality filtering, and data distribution alignment. We integrate existing methods and propose new solutions, such as computing CLIP score on horizontally flipped images to mitigate the interference of scene text, using vision and language models to retrieve training samples for target downstream tasks, rebalancing the data distribution to improve the efficiency of allocating the computational budget, etc. We slice and dice our design choices, provide in-depth analysis, and discuss open questions. Our approach outperforms the best method from the DataComp paper by over 4% on the average performance of 38 tasks and by over 2% on ImageNet.

cs.CV↗
arXiv2023

Dataset Condensation via Generative Model

Dataset condensation aims to condense a large dataset with a lot of training samples into a small set. Previous methods usually condense the dataset into the pixels format. However, it suffers from slow optimization speed and large number of parameters to be optimized. When increasing image resolutions and classes, the number of learnable parameters grows accordingly, prohibiting condensation methods from scaling up to large datasets with diverse classes. Moreover, the relations among condensed samples have been neglected and hence the feature distribution of condensed samples is often not diverse. To solve these problems, we propose to condense the dataset into another format, a generative model. Such a novel format allows for the condensation of large datasets because the size of the generative model remains relatively stable as the number of classes or image resolution increases. Furthermore, an intra-class and an inter-class loss are proposed to model the relation of condensed samples. Intra-class loss aims to create more diverse samples for each class by pushing each sample away from the others of the same class. Meanwhile, inter-class loss increases the discriminability of samples by widening the gap between the centers of different classes. Extensive comparisons with state-of-the-art methods and our ablation studies confirm the effectiveness of our method and its individual component. To our best knowledge, we are the first to successfully conduct condensation on ImageNet-1k.

cs.CV↗
arXiv2023

Why Is Prompt Tuning for Vision-Language Models Robust to Noisy Labels?

Vision-language models such as CLIP learn a generic text-image embedding from large-scale training data. A vision-language model can be adapted to a new classification task through few-shot prompt tuning. We find that such a prompt tuning process is highly robust to label noises. This intrigues us to study the key reasons contributing to the robustness of the prompt tuning paradigm. We conducted extensive experiments to explore this property and find the key factors are: 1) the fixed classname tokens provide a strong regularization to the optimization of the model, reducing gradients induced by the noisy samples; 2) the powerful pre-trained image-text embedding that is learned from diverse and generic web data provides strong prior knowledge for image classification. Further, we demonstrate that noisy zero-shot predictions from CLIP can be used to tune its own prompt, significantly enhancing prediction accuracy in the unsupervised setting. The code is available at https://github.com/CEWu/PTNL.

cs.CV↗
arXiv2023

Exploring the Role of Audio in Video Captioning

Recent focus in video captioning has been on designing architectures that can consume both video and text modalities, and using large-scale video datasets with text transcripts for pre-training, such as HowTo100M. Though these approaches have achieved significant improvement, the audio modality is often ignored in video captioning. In this work, we present an audio-visual framework, which aims to fully exploit the potential of the audio modality for captioning. Instead of relying on text transcripts extracted via automatic speech recognition (ASR), we argue that learning with raw audio signals can be more beneficial, as audio has additional information including acoustic events, speaker identity, etc. Our contributions are twofold. First, we observed that the model overspecializes to the audio modality when pre-training with both video and audio modality, since the ground truth (i.e., text transcripts) can be solely predicted using audio. We proposed a Modality Balanced Pre-training (MBP) loss to mitigate this issue and significantly improve the performance on downstream tasks. Second, we slice and dice different design choices of the cross-modal module, which may become an information bottleneck and generate inferior results. We proposed new local-global fusion mechanisms to improve information exchange across audio and video. We demonstrate significant improvements by leveraging the audio modality on four datasets, and even outperform the state of the art on some metrics without relying on the text modality as the input.

cs.CV↗
arXiv2023

BotMoE: Twitter Bot Detection with Community-Aware Mixtures of Modal-Specific Experts

Twitter bot detection has become a crucial task in efforts to combat online misinformation, mitigate election interference, and curb malicious propaganda. However, advanced Twitter bots often attempt to mimic the characteristics of genuine users through feature manipulation and disguise themselves to fit in diverse user communities, posing challenges for existing Twitter bot detection models. To this end, we propose BotMoE, a Twitter bot detection framework that jointly utilizes multiple user information modalities (metadata, textual content, network structure) to improve the detection of deceptive bots. Furthermore, BotMoE incorporates a community-aware Mixture-of-Experts (MoE) layer to improve domain generalization and adapt to different Twitter communities. Specifically, BotMoE constructs modal-specific encoders for metadata features, textual content, and graphical structure, which jointly model Twitter users from three modal-specific perspectives. We then employ a community-aware MoE layer to automatically assign users to different communities and leverage the corresponding expert networks. Finally, user representations from metadata, text, and graph perspectives are fused with an expert fusion layer, combining all three modalities while measuring the consistency of user information. Extensive experiments demonstrate that BotMoE significantly advances the state-of-the-art on three Twitter bot detection benchmarks. Studies also confirm that BotMoE captures advanced and evasive bots, alleviates the reliance on training data, and better generalizes to new and previously unseen user communities.

cs.SI↗
arXiv2023

$R^{2}$Former: Unified $R$etrieval and $R$eranking Transformer for Place Recognition

Visual Place Recognition (VPR) estimates the location of query images by matching them with images in a reference database. Conventional methods generally adopt aggregated CNN features for global retrieval and RANSAC-based geometric verification for reranking. However, RANSAC only employs geometric information but ignores other possible information that could be useful for reranking, e.g. local feature correlations, and attention values. In this paper, we propose a unified place recognition framework that handles both retrieval and reranking with a novel transformer model, named $R^{2}$Former. The proposed reranking module takes feature correlation, attention value, and xy coordinates into account, and learns to determine whether the image pair is from the same location. The whole pipeline is end-to-end trainable and the reranking module alone can also be adopted on other CNN or transformer backbones as a generic component. Remarkably, $R^{2}$Former significantly outperforms state-of-the-art methods on major VPR datasets with much less inference time and memory consumption. It also achieves the state-of-the-art on the hold-out MSLS challenge set and could serve as a simple yet strong solution for real-world large-scale applications. Experiments also show vision transformer tokens are comparable and sometimes better than CNN local features on local matching. The code is released at https://github.com/Jeff-Zilence/R2Former.

cs.CV↗
arXiv2023

PAniC-3D: Stylized Single-view 3D Reconstruction from Portraits of Anime Characters

We propose PAniC-3D, a system to reconstruct stylized 3D character heads directly from illustrated (p)ortraits of (ani)me (c)haracters. Our anime-style domain poses unique challenges to single-view reconstruction; compared to natural images of human heads, character portrait illustrations have hair and accessories with more complex and diverse geometry, and are shaded with non-photorealistic contour lines. In addition, there is a lack of both 3D model and portrait illustration data suitable to train and evaluate this ambiguous stylized reconstruction task. Facing these challenges, our proposed PAniC-3D architecture crosses the illustration-to-3D domain gap with a line-filling model, and represents sophisticated geometries with a volumetric radiance field. We train our system with two large new datasets (11.2k Vroid 3D models, 1k Vtuber portrait illustrations), and evaluate on a novel AnimeRecon benchmark of illustration-to-3D pairs. PAniC-3D significantly outperforms baseline methods, and provides data to establish the task of stylized reconstruction from portrait illustrations.

cs.CV↗
arXiv2023

Simple and Fast Polarization Tracking algorithm for Continuous-Variable Quantum Key Distribution System Using Orthogonal Pilot Tone

To reduce the influence of random channel polarization variation, especially fast polarization perturbation,for continuous-variable quantum key distribution (CV-QKD) systems, a simple and fast polarization tracking algorithm is proposed and experimentally demonstrated. The algorithm is implemented by an orthogonal pilot tone scheme, one of the pilot tones is used for estimating the polarization rotation angle, and the other one is used for compensating the polarization perturbation introduced phase noise. Meanwhile, residual effects are compensated precisely with the help of a real-valued FIR filter. In this case, the excess noise introduced by polarization perturbation is effectively suppressed. Experimental results show that the polarization scrambling rate 12.57 krad/s can be tracked by using the proposed algorithm, and a good estimated parameters performance is achieved. Compared to conventional polarization tracking algorithms such as the constant modulus algorithm (CMA), experimental results show that the polarization tracking apability of the proposed algorithm is significantly improved. Furthermore, much faster polarization tracking performance is evaluated by digital simulations, and the simulation results show that about 188.50 Mrad/s can be tracked by the proposed algorithm. Thus, our method provides effective technology for the practical application of fiber-based CV-QKD.

quant-ph↗
arXiv2023

Practical Continuous-variable Quantum Key Distribution with Feasible Optimization Parameters

Continuous-variable quantum key distribution (CV-QKD) offers an approach to achieve a potential high secret key rate (SKR) in metropolitan areas. There are several challenges in developing a practical CV-QKD system from the laboratory to the real world. One of the most significant points is that it is really hard to adapt different practical optical fiber conditions for CV-QKD systems with unified hardware. Thus, how to improve the performance of practical CV-QKD systems in the field without modification of the hardware is very important. Here, a systematic optimization method, combining the modulation variance and error correction matrix optimization, is proposed to improve the performance of a practical CV-QKD system with a restricted capacity of postprocessing. The effect of restricted postprocessing capacity on the SKR is modeled as a nonlinear programming problem with modulation variance as an optimization parameter, and the selection of an optimal error correction matrix is studied under the same scheme. The results show that the SKR of a CV-QKD system can be improved by 24% and 200% compared with previous frequently used optimization methods theoretically with a transmission distance of 50 km. Furthermore, the experimental results verify the feasibility and robustness of the proposed method, where the achieved optimal SKR achieved practically deviates <1.6% from the theoretical optimal value. Our results pave the way to deploy high-performance CV-QKD in the real world.

quant-ph↗