arXiv subjects

Hao Tan

Publications and source records attributed to Hao Tan.

At least 91 records · Page 5Linked to original sources

arXiv2024

DocumentCLIP: Linking Figures and Main Body Text in Reflowed Documents

Vision-language pretraining models have achieved great success in supporting multimedia applications by understanding the alignments between images and text. While existing vision-language pretraining models primarily focus on understanding single image associated with a single piece of text, they often ignore the alignment at the intra-document level, consisting of multiple sentences with multiple images. In this work, we propose DocumentCLIP, a salience-aware contrastive learning framework to enforce vision-language pretraining models to comprehend the interaction between images and longer text within documents. Our model is beneficial for the real-world multimodal document understanding like news article, magazines, product descriptions, which contain linguistically and visually richer content. To the best of our knowledge, we are the first to explore multimodal intra-document links by contrastive learning. In addition, we collect a large Wikipedia dataset for pretraining, which provides various topics and structures. Experiments show DocumentCLIP not only outperforms the state-of-the-art baselines in the supervised setting, but also achieves the best zero-shot performance in the wild after human evaluation. Our code is available at https://github.com/FuxiaoLiu/DocumentCLIP.

cs.CV↗

arXiv2024

Carve3D: Improving Multi-view Reconstruction Consistency for Diffusion Models with RL Finetuning

Multi-view diffusion models, obtained by applying Supervised Finetuning (SFT) to text-to-image diffusion models, have driven recent breakthroughs in text-to-3D research. However, due to the limited size and quality of existing 3D datasets, they still suffer from multi-view inconsistencies and Neural Radiance Field (NeRF) reconstruction artifacts. We argue that multi-view diffusion models can benefit from further Reinforcement Learning Finetuning (RLFT), which allows models to learn from the data generated by themselves and improve beyond their dataset limitations during SFT. To this end, we introduce Carve3D, an improved RLFT algorithm coupled with a novel Multi-view Reconstruction Consistency (MRC) metric, to enhance the consistency of multi-view diffusion models. To measure the MRC metric on a set of multi-view images, we compare them with their corresponding NeRF renderings at the same camera viewpoints. The resulting model, which we denote as Carve3DM, demonstrates superior multi-view consistency and NeRF reconstruction quality than existing models. Our results suggest that pairing SFT with Carve3D's RLFT is essential for developing multi-view-consistent diffusion models, mirroring the standard Large Language Model (LLM) alignment pipeline. Our code, training and testing data, and video results are available at: https://desaixie.github.io/carve-3d.

cs.CV↗

arXiv2024

Surface variation analysis of freeform optical systems over surface frequency bands for prescribed wavefront errors

The surface errors of freeform surfaces reflect the manufacturing complexities and significantly impact the feasibility of processing designed optical systems. With multiple degrees of freedom, freeform surfaces pose challenges in surface tolerance analysis in the field. Nevertheless, current research has neglected the influence of surface slopes on the directions of ray propagation. A sudden alteration in the surface slope will lead to a corresponding abrupt shift in the wavefront, even when the change in surface sag is minimal. Moreover, within the realm of freeform surface manufacturing, variation in surface slope across different frequency bands may give rise to unique surface variation. Within the context of this study, we propose a tolerance analysis method to analyze surface variation in freeform surfaces considering surface frequency band slopes based on real ray data. This approach utilizes real ray data to rapidly evaluate surface variation within a specified frequency band of surface slopes. Crucially, our proposed method yields the capability to obtain system surface variation with significant wavefront aberration, in contrast to previous methodologies. The feasibility and advantages of this framework are assessed by analyzing a single-mirror system with a single field and an off-axis two-mirror system. We expect to integrate the proposed methodology with freeform surface design and manufacturing, thereby expanding the scope of freeform optics.

physics.optics↗

arXiv2024

Template-Free Single-View 3D Human Digitalization with Diffusion-Guided LRM

Reconstructing 3D humans from a single image has been extensively investigated. However, existing approaches often fall short on capturing fine geometry and appearance details, hallucinating occluded parts with plausible details, and achieving generalization across unseen and in-the-wild datasets. We present Human-LRM, a diffusion-guided feed-forward model that predicts the implicit field of a human from a single image. Leveraging the power of the state-of-the-art reconstruction model (i.e., LRM) and generative model (i.e Stable Diffusion), our method is able to capture human without any template prior, e.g., SMPL, and effectively enhance occluded parts with rich and realistic details. Our approach first uses a single-view LRM model with an enhanced geometry decoder to get the triplane NeRF representation. The novel view renderings from the triplane NeRF provide strong geometry and color prior, from which we generate photo-realistic details for the occluded parts using a diffusion model. The generated multiple views then enable reconstruction with high-quality geometry and appearance, leading to superior overall performance comparing to all existing human reconstruction methods.

cs.CV↗

arXiv2024

LRM: Large Reconstruction Model for Single Image to 3D

We propose the first Large Reconstruction Model (LRM) that predicts the 3D model of an object from a single input image within just 5 seconds. In contrast to many previous methods that are trained on small-scale datasets such as ShapeNet in a category-specific fashion, LRM adopts a highly scalable transformer-based architecture with 500 million learnable parameters to directly predict a neural radiance field (NeRF) from the input image. We train our model in an end-to-end manner on massive multi-view data containing around 1 million objects, including both synthetic renderings from Objaverse and real captures from MVImgNet. This combination of a high-capacity model and large-scale training data empowers our model to be highly generalizable and produce high-quality 3D reconstructions from various testing inputs, including real-world in-the-wild captures and images created by generative models. Video demos and interactable 3D meshes can be found on our LRM project webpage: https://yiconghong.me/LRM.

cs.CV↗

arXiv2024

PVLR: Prompt-driven Visual-Linguistic Representation Learning for Multi-Label Image Recognition

Multi-label image recognition is a fundamental task in computer vision. Recently, vision-language models have made notable advancements in this area. However, previous methods often failed to effectively leverage the rich knowledge within language models and instead incorporated label semantics into visual features in a unidirectional manner. In this paper, we propose a Prompt-driven Visual-Linguistic Representation Learning (PVLR) framework to better leverage the capabilities of the linguistic modality. In PVLR, we first introduce a dual-prompting strategy comprising Knowledge-Aware Prompting (KAP) and Context-Aware Prompting (CAP). KAP utilizes fixed prompts to capture the intrinsic semantic knowledge and relationships across all labels, while CAP employs learnable prompts to capture context-aware label semantics and relationships. Later, we propose an Interaction and Fusion Module (IFM) to interact and fuse the representations obtained from KAP and CAP. In contrast to the unidirectional fusion in previous works, we introduce a Dual-Modal Attention (DMA) that enables bidirectional interaction between textual and visual features, yielding context-aware label representations and semantic-related visual representations, which are subsequently used to calculate similarities and generate final predictions for all labels. Extensive experiments on three popular datasets including MS-COCO, Pascal VOC 2007, and NUS-WIDE demonstrate the superiority of PVLR.

cs.CV↗

arXiv2024

Effect of light injection on the security of practical quantum key distribution

Quantum key distribution (QKD) based on the fundamental laws of quantum physics can allow the distribution of secure keys between distant users. However, the imperfections in realistic devices may lead to potential security risks, which must be accurately characterized and considered in practical security analysis. High-speed optical modulators, being as one of the core components of practical QKD systems, can be used to prepare the required quantum states. Here, we find that optical modulators based on LiNbO3, including phase modulators and intensity modulators, are vulnerable to photorefractive effect caused by external light injection. By changing the power of external light, eavesdroppers can control the intensities of the prepared states, posing a potential threat to the security of QKD. We have experimentally demonstrated the influence of light injection on LiNbO3-based optical modulators and analyzed the security risks caused by the potential green light injection attack, along with the corresponding countermeasures.

quant-ph↗

arXiv2023

Compound Text-Guided Prompt Tuning via Image-Adaptive Cues

Vision-Language Models (VLMs) such as CLIP have demonstrated remarkable generalization capabilities to downstream tasks. However, existing prompt tuning based frameworks need to parallelize learnable textual inputs for all categories, suffering from massive GPU memory consumption when there is a large number of categories in the target dataset. Moreover, previous works require to include category names within prompts, exhibiting subpar performance when dealing with ambiguous category names. To address these shortcomings, we propose Compound Text-Guided Prompt Tuning (TGP-T) that significantly reduces resource demand while achieving superior performance. We introduce text supervision to the optimization of prompts, which enables two benefits: 1) releasing the model reliance on the pre-defined category names during inference, thereby enabling more flexible prompt generation; 2) reducing the number of inputs to the text encoder, which decreases GPU memory consumption significantly. Specifically, we found that compound text supervisions, i.e., category-wise and content-wise, is highly effective, since they provide inter-class separability and capture intra-class variations, respectively. Moreover, we condition the prompt generation on visual features through a module called Bonder, which facilitates the alignment between prompts and visual features. Extensive experiments on few-shot recognition and domain generalization demonstrate that TGP-T achieves superior performance with consistently lower training costs. It reduces GPU memory usage by 93% and attains a 2.5% performance gain on 16-shot ImageNet. The code is available at https://github.com/EricTan7/TGP-T.

cs.CV↗

arXiv2023

Instant3D: Fast Text-to-3D with Sparse-View Generation and Large Reconstruction Model

Text-to-3D with diffusion models has achieved remarkable progress in recent years. However, existing methods either rely on score distillation-based optimization which suffer from slow inference, low diversity and Janus problems, or are feed-forward methods that generate low-quality results due to the scarcity of 3D training data. In this paper, we propose Instant3D, a novel method that generates high-quality and diverse 3D assets from text prompts in a feed-forward manner. We adopt a two-stage paradigm, which first generates a sparse set of four structured and consistent views from text in one shot with a fine-tuned 2D text-to-image diffusion model, and then directly regresses the NeRF from the generated images with a novel transformer-based sparse-view reconstructor. Through extensive experiments, we demonstrate that our method can generate diverse 3D assets of high visual quality within 20 seconds, which is two orders of magnitude faster than previous optimization-based methods that can take 1 to 10 hours. Our project webpage: https://jiahao.ai/instant3d/.

cs.CV↗

arXiv2023

PF-LRM: Pose-Free Large Reconstruction Model for Joint Pose and Shape Prediction

We propose a Pose-Free Large Reconstruction Model (PF-LRM) for reconstructing a 3D object from a few unposed images even with little visual overlap, while simultaneously estimating the relative camera poses in ~1.3 seconds on a single A100 GPU. PF-LRM is a highly scalable method utilizing the self-attention blocks to exchange information between 3D object tokens and 2D image tokens; we predict a coarse point cloud for each view, and then use a differentiable Perspective-n-Point (PnP) solver to obtain camera poses. When trained on a huge amount of multi-view posed data of ~1M objects, PF-LRM shows strong cross-dataset generalization ability, and outperforms baseline methods by a large margin in terms of pose prediction accuracy and 3D reconstruction quality on various unseen evaluation datasets. We also demonstrate our model's applicability in downstream text/image-to-3D task with fast feed-forward inference. Our project website is at: https://totoro97.github.io/pf-lrm .

cs.CV↗

arXiv2023

DMV3D: Denoising Multi-View Diffusion using 3D Large Reconstruction Model

We propose \textbf{DMV3D}, a novel 3D generation approach that uses a transformer-based 3D large reconstruction model to denoise multi-view diffusion. Our reconstruction model incorporates a triplane NeRF representation and can denoise noisy multi-view images via NeRF reconstruction and rendering, achieving single-stage 3D generation in $\sim$30s on single A100 GPU. We train \textbf{DMV3D} on large-scale multi-view image datasets of highly diverse objects using only image reconstruction losses, without accessing 3D assets. We demonstrate state-of-the-art results for the single-image reconstruction problem where probabilistic modeling of unseen object parts is required for generating diverse reconstructions with sharp textures. We also show high-quality text-to-3D generation results outperforming previous 3D diffusion models. Our project website is at: https://justimyhxu.github.io/projects/dmv3d/ .

cs.CV↗

arXiv2023

Federated Skewed Label Learning with Logits Fusion

Federated learning (FL) aims to collaboratively train a shared model across multiple clients without transmitting their local data. Data heterogeneity is a critical challenge in realistic FL settings, as it causes significant performance deterioration due to discrepancies in optimization among local models. In this work, we focus on label distribution skew, a common scenario in data heterogeneity, where the data label categories are imbalanced on each client. To address this issue, we propose FedBalance, which corrects the optimization bias among local models by calibrating their logits. Specifically, we introduce an extra private weak learner on the client side, which forms an ensemble model with the local model. By fusing the logits of the two models, the private weak learner can capture the variance of different data, regardless of their category. Therefore, the optimization direction of local models can be improved by increasing the penalty for misclassifying minority classes and reducing the attention to majority classes, resulting in a better global model. Extensive experiments show that our method can gain 13\% higher average accuracy compared with state-of-the-art methods.

cs.LG↗

arXiv2023

Scaling Data Generation in Vision-and-Language Navigation

Recent research in language-guided visual navigation has demonstrated a significant demand for the diversity of traversable environments and the quantity of supervision for training generalizable agents. To tackle the common data scarcity issue in existing vision-and-language navigation datasets, we propose an effective paradigm for generating large-scale data for learning, which applies 1200+ photo-realistic environments from HM3D and Gibson datasets and synthesizes 4.9 million instruction trajectory pairs using fully-accessible resources on the web. Importantly, we investigate the influence of each component in this paradigm on the agent's performance and study how to adequately apply the augmented data to pre-train and fine-tune an agent. Thanks to our large-scale dataset, the performance of an existing agent can be pushed up (+11% absolute with regard to previous SoTA) to a significantly new best of 80% single-run success rate on the R2R test split by simple imitation learning. The long-lasting generalization gap between navigating in seen and unseen environments is also reduced to less than 1% (versus 8% in the previous best method). Moreover, our paradigm also facilitates different models to achieve new state-of-the-art navigation results on CVDN, REVERIE, and R2R in continuous environments.

cs.CV↗

arXiv2023

Boosting Punctuation Restoration with Data Generation and Reinforcement Learning

Punctuation restoration is an important task in automatic speech recognition (ASR) which aim to restore the syntactic structure of generated ASR texts to improve readability. While punctuated texts are abundant from written documents, the discrepancy between written punctuated texts and ASR texts limits the usability of written texts in training punctuation restoration systems for ASR texts. This paper proposes a reinforcement learning method to exploit in-topic written texts and recent advances in large pre-trained generative language models to bridge this gap. The experiments show that our method achieves state-of-the-art performance on the ASR test set on two benchmark datasets for punctuation restoration.

cs.CL↗

arXiv2023

Learning Navigational Visual Representations with Semantic Map Supervision

Being able to perceive the semantics and the spatial structure of the environment is essential for visual navigation of a household robot. However, most existing works only employ visual backbones pre-trained either with independent images for classification or with self-supervised learning methods to adapt to the indoor navigation domain, neglecting the spatial relationships that are essential to the learning of navigation. Inspired by the behavior that humans naturally build semantically and spatially meaningful cognitive maps in their brains during navigation, in this paper, we propose a novel navigational-specific visual representation learning method by contrasting the agent's egocentric views and semantic maps (Ego$^2$-Map). We apply the visual transformer as the backbone encoder and train the model with data collected from the large-scale Habitat-Matterport3D environments. Ego$^2$-Map learning transfers the compact and rich information from a map, such as objects, structure and transition, to the agent's egocentric representations for navigation. Experiments show that agents using our learned representations on object-goal navigation outperform recent visual pre-training methods. Moreover, our representations significantly improve vision-and-language navigation in continuous environments for both high-level and low-level action spaces, achieving new state-of-the-art results of 47% SR and 41% SPL on the test server.

cs.CV↗

arXiv2023

High-rate quantum key distribution exceeding 110 Mb/s

Quantum key distribution (QKD) can provide fundamentally proven security for secure communication. Toward application, the secret key rate (SKR) is a key figure of merit for any QKD system. So far, the SKR has been limited to about a few megabit-per-second. Here we report a QKD system that is able to generate key at a record high SKR of 115.8 Mb/s over 10-km standard fibre, and to distribute key over up to 328 km of ultra-low-loss fibre. This attributes to a multi-pixel superconducting nanowire single-photon detector with ultrahigh counting rate, an integrated transmitter that can stably encode polarization states with low error, a fast post-processing algorithm for generating key in real time and the high system clock-rate operation. The results demonstrate the feasibility of practical high-rate QKD with photonic techniques, thus opening its possibility for widespread applications.

quant-ph↗

arXiv2023

Pupil aberrations correction of the afocal telescope for the TianQin project

TianQin is a planned Chinese space-based gravitational wave (GW) observatory with a frequency band of 10-4 to 1Hz. Optical telescopes are essential for the delivery of the measurement beam to support a precise distance measurement between pairs of proof masses. As the design is driven by the interferometric displacement sensitivity requirements, the stability control of optical path length (OPL) is extremely important beyond the traditional requirement of diffraction-limited imaging quality. In a telescope system, the recurring tilt-to-length (TTL) coupling noise arises from the OPL variation due to the wavefront deformation and angular misalignment. The pupil aberrations are preferred option to understand the OPL specifications and further suppress TTL coupling noise. To correct the pupil aberrations, we derive primary pupil aberrations in a series expansion form, and then refine the formulation of merit function by combining the pupil aberration theory and traditional image aberration theory. The automatic correction of pupil aberrations is carried out by using the macro programming in the commercial optical software Zemax, leading to a high performance telescope design. The design results show that on one side the pupil aberrations have been corrected, and on the other side, its optical performance meets the requirements for TianQin project. The RMS wavefront error over the science field of view (FOV) is less than λ/200 and the maximum TTL coupling noise over the entire 300 urad FOV is 0.0034nm/urad. We believe that our design approach can be a good guide for the space telescope design in any other space-based GW detection project, as well as other similar optical systems.

astro-ph.IM↗

arXiv2022

Tiny-Attention Adapter: Contexts Are More Important Than the Number of Parameters

Adapter-tuning is a paradigm that transfers a pretrained language model to downstream tasks by adding and tuning a small number of new parameters. Previously proposed adapter architectures are all feed-forward neural networks. In this paper, we investigate the effectiveness of using tiny-attention -- i.e., attention with extremely small per-head dimensionality -- as adapters. Our tiny-attention adapter learns to modify the hidden states at each position directly conditioned on the hidden states at all the other positions, which is missed by the previously proposed adapters. Moreover, we view its multiple attention heads as a mixture of experts and propose to average their weights during deployment, which further reduces its inference computation cost. On the GLUE benchmark, our tiny-attention adapter outperforms the other parameter-efficient transfer learning methods as well as full fine-tuning while only updating 0.05% of the parameters. On the FewGLUE benchmark, its performance is comparable to that of GPT-3 and PET.

cs.CL↗