arXiv subjects

Han Zhang

Publications and source records attributed to Han Zhang.

At least 325 records · Page 18Linked to original sources

arXiv2023

NEON: Living Needs Prediction System in Meituan

Living needs refer to the various needs in human's daily lives for survival and well-being, including food, housing, entertainment, etc. On life service platforms that connect users to service providers, such as Meituan, the problem of living needs prediction is fundamental as it helps understand users and boost various downstream applications such as personalized recommendation. However, the problem has not been well explored and is faced with two critical challenges. First, the needs are naturally connected to specific locations and times, suffering from complex impacts from the spatiotemporal context. Second, there is a significant gap between users' actual living needs and their historical records on the platform. To address these two challenges, we design a system of living NEeds predictiON named NEON, consisting of three phases: feature mining, feature fusion, and multi-task prediction. In the feature mining phase, we carefully extract individual-level user features for spatiotemporal modeling, and aggregated-level behavioral features for enriching data, which serve as the basis for addressing two challenges, respectively. Further, in the feature fusion phase, we propose a neural network that effectively fuses two parts of features into the user representation. Moreover, we design a multi-task prediction phase, where the auxiliary task of needs-meeting way prediction can enhance the modeling of spatiotemporal context. Extensive offline evaluations verify that our NEON system can effectively predict users' living needs. Furthermore, we deploy NEON into Meituan's algorithm engine and evaluate how it enhances the three downstream prediction applications, via large-scale online A/B testing.

cs.IR↗

arXiv2023

Space-Air-Ground Integrated Network (SAGIN): A Survey

Since existing mobile communication networks may not be able to meet the low latency and high-efficiency requirements of emerging technologies and applications, novel network architectures need to be investigated to support these new requirements. As a new network architecture that integrates satellite systems, air networks and ground communication, Space-Air-Ground Integrated Network (SAGIN) has attracted extensive attention in recent years. This paper summarizes the recent research work on SAGIN from several aspects, with the basic information of SAGIN first introduced, followed by the physical characteristics. Then the drive and prospects of the current SAGIN architecture in supporting new requirements are deeply analyzed. On this basis, the requirements and challenges are analyzed. Finally, it summarizes the existing solutions and prospects the future research directions.

cs.NI↗

arXiv2023

SVDiff: Compact Parameter Space for Diffusion Fine-Tuning

Diffusion models have achieved remarkable success in text-to-image generation, enabling the creation of high-quality images from text prompts or other modalities. However, existing methods for customizing these models are limited by handling multiple personalized subjects and the risk of overfitting. Moreover, their large number of parameters is inefficient for model storage. In this paper, we propose a novel approach to address these limitations in existing text-to-image diffusion models for personalization. Our method involves fine-tuning the singular values of the weight matrices, leading to a compact and efficient parameter space that reduces the risk of overfitting and language drifting. We also propose a Cut-Mix-Unmix data-augmentation technique to enhance the quality of multi-subject image generation and a simple text-based image editing framework. Our proposed SVDiff method has a significantly smaller model size compared to existing methods (approximately 2,200 times fewer parameters compared with vanilla DreamBooth), making it more practical for real-world applications.

cs.CV↗

arXiv2023

MAGE: MAsked Generative Encoder to Unify Representation Learning and Image Synthesis

Generative modeling and representation learning are two key tasks in computer vision. However, these models are typically trained independently, which ignores the potential for each task to help the other, and leads to training and model maintenance overheads. In this work, we propose MAsked Generative Encoder (MAGE), the first framework to unify SOTA image generation and self-supervised representation learning. Our key insight is that using variable masking ratios in masked image modeling pre-training can allow generative training (very high masking ratio) and representation learning (lower masking ratio) under the same training framework. Inspired by previous generative models, MAGE uses semantic tokens learned by a vector-quantized GAN at inputs and outputs, combining this with masking. We can further improve the representation by adding a contrastive loss to the encoder output. We extensively evaluate the generation and representation learning capabilities of MAGE. On ImageNet-1K, a single MAGE ViT-L model obtains 9.10 FID in the task of class-unconditional image generation and 78.9% top-1 accuracy for linear probing, achieving state-of-the-art performance in both image generation and representation learning. Code is available at https://github.com/LTH14/mage.

cs.CV↗

arXiv2023

Legendrian Negative Torus Knots in Universally Tight Lens Spaces

The main theorem characterizes all Legendrian negative torus knots in universally tight lens space in the sense of coarse equivalence. Together with Onaran's results on Legendrian positive torus knots, all Legendrian torus knots in universally tight lens space is classified. The main method is splitting L(p, q) by convex Heegaard decomposition.

math.GT↗

arXiv2023

Learning Disentangled Prompts for Compositional Image Synthesis

We study domain-adaptive image synthesis, the problem of teaching pretrained image generative models a new style or concept from as few as one image to synthesize novel images, to better understand the compositional image synthesis. We present a framework that leverages a pretrained class-conditional generation model and visual prompt tuning. Specifically, we propose a novel source class distilled visual prompt that learns disentangled prompts of semantic (e.g., class) and domain (e.g., style) from a few images. Learned domain prompt is then used to synthesize images of any classes in the style of target domain. We conduct studies on various target domains with the number of images ranging from one to a few to many, and show qualitative results which show the compositional generalization of our method. Moreover, we show that our method can help improve zero-shot domain adaptation classification accuracy.

cs.CV↗

arXiv2023

Inverse optimal control for averaged cost per stage linear quadratic regulators

Inverse Optimal Control (IOC) is a powerful framework for learning a behaviour from observations of experts. The framework aims to identify the underlying cost function that the observed optimal trajectories (the experts' behaviour) are optimal with respect to. In this work, we considered the case of identifying the cost and the feedback law from observed trajectories generated by an ``average cost per stage" linear quadratic regulator. We show that identifying the cost is in general an ill-posed problem, and give necessary and sufficient conditions for non-identifiability. Moreover, despite the fact that the problem is in general ill-posed, we construct an estimator for the cost function and show that the control gain corresponding to this estimator is a statistically consistent estimator for the true underlying control gain. In fact, the constructed estimator is based on convex optimization, and hence the proved statistical consistency is also observed in practice. We illustrate the latter by applying the method on a simulation example from rehabilitation robotics.

math.OC↗

arXiv2023

Enhancing the Performance of Transformer-based Spiking Neural Networks by SNN-optimized Downsampling with Precise Gradient Backpropagation

Deep spiking neural networks (SNNs) have drawn much attention in recent years because of their low power consumption, biological rationality and event-driven property. However, state-of-the-art deep SNNs (including Spikformer and Spikingformer) suffer from a critical challenge related to the imprecise gradient backpropagation. This problem arises from the improper design of downsampling modules in these networks, and greatly hampering the overall model performance. In this paper, we propose ConvBN-MaxPooling-LIF (CML), an SNN-optimized downsampling with precise gradient backpropagation. We prove that CML can effectively overcome the imprecision of gradient backpropagation from a theoretical perspective. In addition, we evaluate CML on ImageNet, CIFAR10, CIFAR100, CIFAR10-DVS, DVS128-Gesture datasets, and show state-of-the-art performance on all these datasets with significantly enhanced performances compared with Spikingformer. For instance, our model achieves 77.64 $\%$ on ImageNet, 96.04 $\%$ on CIFAR10, 81.4$\%$ on CIFAR10-DVS, with + 1.79$\%$ on ImageNet, +1.16$\%$ on CIFAR100 compared with Spikingformer.

cs.NE↗

arXiv2023

Investigating $η^\prime_{1}(1855)$ exotic states in $J/ψ\toη^\prime_{1}(1855)η^{(\prime)}$ decays

An analysis of the $J/ψ\to ηη^\primeγ$ decay by the BESIII collaboration claims the observation of an exotic state $η_1(1855)$ with $I^GJ^{PC}=0^+1^{-+}$. To establish its C-parity partner $η^\prime_{1}(1855)$ in the picture of the $K \bar{K}_1(1400)$ molecular state, we propose that $J/ψ\toη^\prime_{1}(1855)η^{(\prime)}$ receives the main contributions from the final state interaction of $KK^*$($K^+ K^{*-}$, $K^- K^{*+}$, $K^0 \bar{K}^{*0}$, and $\bar{K}^0 K^{*0}$). Specifically, $K$ and $K^*$ in $J/ψ\to KK^*$ decays transform as $η^\prime_{1}(1855)η^{(\prime)}$, by exchanging a $K_1(1400)$. We predict ${\cal B}(J/ψ\to η^{\prime}_1(1855)η)=(6.3^{+12.6}_{-3.5})\times 10^{-6}$, and ${\cal B}(J/ψ\to η^{\prime}_1(1855)η^{\prime}) =(6.5^{+6.6}_{-4.6})\times 10^{-6}$, which can be studied in the $J/ψ\to K^*\bar{K}^*η^{(\prime)}$ decays.

hep-ph↗

arXiv2023

Mining fMRI Dynamics with Parcellation Prior for Brain Disease Diagnosis

To characterize atypical brain dynamics under diseases, prevalent studies investigate functional magnetic resonance imaging (fMRI). However, most of the existing analyses compress rich spatial-temporal information as the brain functional networks (BFNs) and directly investigate the whole-brain network without neurological priors about functional subnetworks. We thus propose a novel graph learning framework to mine fMRI signals with topological priors from brain parcellation for disease diagnosis. Specifically, we 1) detect diagnosis-related temporal features using a "Transformer" for a higher-level BFN construction, and process it with a following graph convolutional network, and 2) apply an attention-based multiple instance learning strategy to emphasize the disease-affected subnetworks to further enhance the diagnosis performance and interpretability. Experiments demonstrate higher effectiveness of our method than compared methods in the diagnosis of early mild cognitive impairment. More importantly, our method is capable of localizing crucial brain subnetworks during the diagnosis, providing insights into the pathogenic source of mild cognitive impairment.

eess.IV↗

arXiv2023

Using Language to Extend to Unseen Domains

It is expensive to collect training data for every possible domain that a vision model may encounter when deployed. We instead consider how simply verbalizing the training domain (e.g. "photos of birds") as well as domains we want to extend to but do not have data for (e.g. "paintings of birds") can improve robustness. Using a multimodal model with a joint image and language embedding space, our method LADS learns a transformation of the image embeddings from the training domain to each unseen test domain, while preserving task relevant information. Without using any images from the unseen test domain, we show that over the extended domain containing both training and unseen test domains, LADS outperforms standard fine-tuning and ensemble approaches over a suite of four benchmarks targeting domain adaptation and dataset bias.

cs.CV↗

arXiv2023

Pelphix: Surgical Phase Recognition from X-ray Images in Percutaneous Pelvic Fixation

Surgical phase recognition (SPR) is a crucial element in the digital transformation of the modern operating theater. While SPR based on video sources is well-established, incorporation of interventional X-ray sequences has not yet been explored. This paper presents Pelphix, a first approach to SPR for X-ray-guided percutaneous pelvic fracture fixation, which models the procedure at four levels of granularity -- corridor, activity, view, and frame value -- simulating the pelvic fracture fixation workflow as a Markov process to provide fully annotated training data. Using added supervision from detection of bony corridors, tools, and anatomy, we learn image representations that are fed into a transformer model to regress surgical phases at the four granularity levels. Our approach demonstrates the feasibility of X-ray-based SPR, achieving an average accuracy of 93.8% on simulated sequences and 67.57% in cadaver across all granularity levels, with up to 88% accuracy for the target corridor in real data. This work constitutes the first step toward SPR for the X-ray domain, establishing an approach to categorizing phases in X-ray-guided surgery, simulating realistic image sequences to enable machine learning model development, and demonstrating that this approach is feasible for the analysis of real procedures. As X-ray-based SPR continues to mature, it will benefit procedures in orthopedic surgery, angiography, and interventional radiology by equipping intelligent surgical systems with situational awareness in the operating room.

cs.LG↗

arXiv2023

Identity Encoder for Personalized Diffusion

Many applications can benefit from personalized image generation models, including image enhancement, video conferences, just to name a few. Existing works achieved personalization by fine-tuning one model for each person. While being successful, this approach incurs additional computation and storage overhead for each new identity. Furthermore, it usually expects tens or hundreds of examples per identity to achieve the best performance. To overcome these challenges, we propose an encoder-based approach for personalization. We learn an identity encoder which can extract an identity representation from a set of reference images of a subject, together with a diffusion generator that can generate new images of the subject conditioned on the identity representation. Once being trained, the model can be used to generate images of arbitrary identities given a few examples even if the model hasn't been trained on the identity. Our approach greatly reduces the overhead for personalized image generation and is more applicable in many potential applications. Empirical results show that our approach consistently outperforms existing fine-tuning based approach in both image generation and reconstruction, and the outputs is preferred by users more than 95% of the time compared with the best performing baseline.

cs.CV↗

arXiv2023

MAGVIT: Masked Generative Video Transformer

We introduce the MAsked Generative VIdeo Transformer, MAGVIT, to tackle various video synthesis tasks with a single model. We introduce a 3D tokenizer to quantize a video into spatial-temporal visual tokens and propose an embedding method for masked video token modeling to facilitate multi-task learning. We conduct extensive experiments to demonstrate the quality, efficiency, and flexibility of MAGVIT. Our experiments show that (i) MAGVIT performs favorably against state-of-the-art approaches and establishes the best-published FVD on three video generation benchmarks, including the challenging Kinetics-600. (ii) MAGVIT outperforms existing methods in inference time by two orders of magnitude against diffusion models and by 60x against autoregressive models. (iii) A single MAGVIT model supports ten diverse generation tasks and generalizes across videos from different visual domains. The source code and trained models will be released to the public at https://magvit.cs.cmu.edu.

cs.CV↗

arXiv2023

Taming Encoder for Zero Fine-tuning Image Customization with Text-to-Image Diffusion Models

This paper proposes a method for generating images of customized objects specified by users. The method is based on a general framework that bypasses the lengthy optimization required by previous approaches, which often employ a per-object optimization paradigm. Our framework adopts an encoder to capture high-level identifiable semantics of objects, producing an object-specific embedding with only a single feed-forward pass. The acquired object embedding is then passed to a text-to-image synthesis model for subsequent generation. To effectively blend a object-aware embedding space into a well developed text-to-image model under the same generation context, we investigate different network designs and training strategies, and propose a simple yet effective regularized joint training scheme with an object identity preservation loss. Additionally, we propose a caption generation scheme that become a critical piece in fostering object specific embedding faithfully reflected into the generation process, while keeping control and editing abilities. Once trained, the network is able to produce diverse content and styles, conditioned on both texts and objects. We demonstrate through experiments that our proposed method is able to synthesize images with compelling output quality, appearance diversity, and object fidelity, without the need of test-time optimization. Systematic studies are also conducted to analyze our models, providing insights for future work.

cs.CV↗

arXiv2023

Enhanced Training of Query-Based Object Detection via Selective Query Recollection

This paper investigates a phenomenon where query-based object detectors mispredict at the last decoding stage while predicting correctly at an intermediate stage. We review the training process and attribute the overlooked phenomenon to two limitations: lack of training emphasis and cascading errors from decoding sequence. We design and present Selective Query Recollection (SQR), a simple and effective training strategy for query-based object detectors. It cumulatively collects intermediate queries as decoding stages go deeper and selectively forwards the queries to the downstream stages aside from the sequential structure. Such-wise, SQR places training emphasis on later stages and allows later stages to work with intermediate queries from earlier stages directly. SQR can be easily plugged into various query-based object detectors and significantly enhances their performance while leaving the inference pipeline unchanged. As a result, we apply SQR on Adamixer, DAB-DETR, and Deformable-DETR across various settings (backbone, number of queries, schedule) and consistently brings 1.4-2.8 AP improvement.

cs.CV↗

arXiv2023

Delayed and Indirect Impacts of Link Recommendations

The impacts of link recommendations on social networks are challenging to evaluate, and so far they have been studied in limited settings. Observational studies are restricted in the kinds of causal questions they can answer and naive A/B tests often lead to biased evaluations due to unaccounted network interference. Furthermore, evaluations in simulation settings are often limited to static network models that do not take into account the potential feedback loops between link recommendation and organic network evolution. To this end, we study the impacts of recommendations on social networks in dynamic settings. Adopting a simulation-based approach, we consider an explicit dynamic formation model -- an extension of the celebrated Jackson-Rogers model -- and investigate how link recommendations affect network evolution over time. Empirically, we find that link recommendations have surprising delayed and indirect effects on the structural properties of networks. Specifically, we find that link recommendations can exhibit considerably different impacts in the immediate term and in the long term. For instance, we observe that friend-of-friend recommendations can have an immediate effect in decreasing degree inequality, but in the long term, they can make the degree distribution substantially more unequal. Moreover, we show that the effects of recommendations can persist in networks, in part due to their indirect impacts on natural dynamics even after recommendations are turned off. We show that, in counterfactual simulations, removing the indirect effects of link recommendations can make the network trend faster toward what it would have been under natural growth dynamics.

cs.SI↗

arXiv2023

Streaming Kernel PCA Algorithm With Small Space

Principal Component Analysis (PCA) is a widely used technique in machine learning, data analysis and signal processing. With the increase in the size and complexity of datasets, it has become important to develop low-space usage algorithms for PCA. Streaming PCA has gained significant attention in recent years, as it can handle large datasets efficiently. The kernel method, which is commonly used in learning algorithms such as Support Vector Machines (SVMs), has also been applied in PCA algorithms. We propose a streaming algorithm for Kernel PCA problems based on the traditional scheme by Oja. Our algorithm addresses the challenge of reducing the memory usage of PCA while maintaining its accuracy. We analyze the performance of our algorithm by studying the conditions under which it succeeds. Specifically, we show that, when the spectral ratio $R := λ_1/λ_2$ of the target covariance matrix is lower bounded by $C \cdot \log n\cdot \log d$, the streaming PCA can be solved with $O(d)$ space cost. Our proposed algorithm has several advantages over existing methods. First, it is a streaming algorithm that can handle large datasets efficiently. Second, it employs the kernel method, which allows it to capture complex nonlinear relationships among data points. Third, it has a low-space usage, making it suitable for applications where memory is limited.

cs.LG↗