Search

Multimodal large language models (MLLMs) have experienced significant advancements recently, but still struggle to recognize and interpret intricate details in high-resolution (HR) images effectively. While state-of-the-art (SOTA) MLLMs claim to process images at 4K resolution, existing MLLM benchmarks only support up to 2K, leaving the capabilities of SOTA models on true HR images largely untested. Furthermore, existing methods for enhancing HR image perception in MLLMs rely on computationally expensive visual instruction tuning. To address these limitations, we introduce HR-Bench, the first deliberately designed benchmark to rigorously evaluate MLLM performance on 4K&8K images. Through extensive experiments, we demonstrate that while downsampling HR images leads to vision information loss, leveraging complementary modalities, e.g., text, can effectively compensate for this loss. Building upon this insight, we propose Divide, Conquer and Combine (DC$^2$), a novel training-free framework for enhancing MLLM perception of HR images. DC$^2$ follows a three-staged approach: 1) Divide: recursively partitioning the HR image into patches and merging similar patches to minimize computational overhead, 2) Conquer: leveraging the MLLM to generate accurate textual descriptions for each image patch, and 3) Combine: utilizing the generated text descriptions to enhance the MLLM's understanding of the overall HR image. Extensive experiments show that: 1) the SOTA MLLM achieves 63% accuracy, which is markedly lower than the 87% accuracy achieved by humans on HR-Bench; 2) our DC$^2$ brings consistent and significant improvements (a relative increase of +6% on HR-Bench and +8% on general multimodal benchmarks). The benchmark and code will be released to facilitate the multimodal R&D community.

Scene Graph Generation (SGG) aims to explore the relationships between objects in images and obtain scene summary graphs, thereby better serving downstream tasks. However, the long-tailed problem has adversely affected the scene graph's quality. The predictions are dominated by coarse-grained relationships, lacking more informative fine-grained ones. The union region of one object pair (i.e., one sample) contains rich and dedicated contextual information, enabling the prediction of the sample-specific bias for refining the original relationship prediction. Therefore, we propose a novel Sample-Level Bias Prediction (SBP) method for fine-grained SGG (SBG). Firstly, we train a classic SGG model and construct a correction bias set by calculating the margin between the ground truth label and the predicted label with one classic SGG model. Then, we devise a Bias-Oriented Generative Adversarial Network (BGAN) that learns to predict the constructed correction biases, which can be utilized to correct the original predictions from coarse-grained relationships to fine-grained ones. The extensive experimental results on VG, GQA, and VG-1800 datasets demonstrate that our SBG outperforms the state-of-the-art methods in terms of Average@K across three mainstream SGG models: Motif, VCtree, and Transformer. Compared to dataset-level correction methods on VG, SBG shows a significant average improvement of 5.6%, 3.9%, and 3.2% on Average@K for tasks PredCls, SGCls, and SGDet, respectively. The code will be available at https://github.com/Zhuzi24/SBG., Comment: 24 pages, 10 figures, ECCV2024

This paper introduces GLOBE, a high-quality English corpus with worldwide accents, specifically designed to address the limitations of current zero-shot speaker adaptive Text-to-Speech (TTS) systems that exhibit poor generalizability in adapting to speakers with accents. Compared to commonly used English corpora, such as LibriTTS and VCTK, GLOBE is unique in its inclusion of utterances from 23,519 speakers and covers 164 accents worldwide, along with detailed metadata for these speakers. Compared to its original corpus, i.e., Common Voice, GLOBE significantly improves the quality of the speech data through rigorous filtering and enhancement processes, while also populating all missing speaker metadata. The final curated GLOBE corpus includes 535 hours of speech data at a 24 kHz sampling rate. Our benchmark results indicate that the speaker adaptive TTS model trained on the GLOBE corpus can synthesize speech with better speaker similarity and comparable naturalness than that trained on other popular corpora. We will release GLOBE publicly after acceptance. The GLOBE dataset is available at https://globecorpus.github.io/., Comment: Interspeech 2024, 4 pages, 3 figures

Scene graph generation (SGG) in satellite imagery (SAI) benefits promoting understanding of geospatial scenarios from perception to cognition. In SAI, objects exhibit great variations in scales and aspect ratios, and there exist rich relationships between objects (even between spatially disjoint objects), which makes it attractive to holistically conduct SGG in large-size very-high-resolution (VHR) SAI. However, there lack such SGG datasets. Due to the complexity of large-size SAI, mining triplets heavily relies on long-range contextual reasoning. Consequently, SGG models designed for small-size natural imagery are not directly applicable to large-size SAI. This paper constructs a large-scale dataset for SGG in large-size VHR SAI with image sizes ranging from 512 x 768 to 27,860 x 31,096 pixels, named STAR (Scene graph generaTion in lArge-size satellite imageRy), encompassing over 210K objects and over 400K triplets. To realize SGG in large-size SAI, we propose a context-aware cascade cognition (CAC) framework to understand SAI regarding object detection (OBD), pair pruning and relationship prediction for SGG. We also release a SAI-oriented SGG toolkit with about 30 OBD and 10 SGG methods which need further adaptation by our devised modules on our challenging STAR dataset. The dataset and toolkit are available at: https://linlin-dev.github.io/project/STAR., Comment: 18 pages, 11 figures

Conventional text-to-speech (TTS) research has predominantly focused on enhancing the quality of synthesized speech for speakers in the training dataset. The challenge of synthesizing lifelike speech for unseen, out-of-dataset speakers, especially those with limited reference data, remains a significant and unresolved problem. While zero-shot or few-shot speaker-adaptive TTS approaches have been explored, they have many limitations. Zero-shot approaches tend to suffer from insufficient generalization performance to reproduce the voice of speakers with heavy accents. While few-shot methods can reproduce highly varying accents, they bring a significant storage burden and the risk of overfitting and catastrophic forgetting. In addition, prior approaches only provide either zero-shot or few-shot adaptation, constraining their utility across varied real-world scenarios with different demands. Besides, most current evaluations of speaker-adaptive TTS are conducted only on datasets of native speakers, inadvertently neglecting a vast portion of non-native speakers with diverse accents. Our proposed framework unifies both zero-shot and few-shot speaker adaptation strategies, which we term as "instant" and "fine-grained" adaptations based on their merits. To alleviate the insufficient generalization performance observed in zero-shot speaker adaptation, we designed two innovative discriminators and introduced a memory mechanism for the speech decoder. To prevent catastrophic forgetting and reduce storage implications for few-shot speaker adaptation, we designed two adapters and a unique adaptation procedure., Comment: 15 pages, 13 figures. Copyright has been transferred to IEEE

Online feedback optimization (OFO) enables optimal steady-state operations of a physical system by employing an iterative optimization algorithm as a dynamic feedback controller. When the plant consists of several interconnected sub-systems, centralized implementations become impractical due to the heavy computational burden and the need to pre-compute system-wide sensitivities, which may not be easily accessible in practice. Motivated by these challenges, we develop a fully distributed model-free OFO controller, featuring consensus-based tracking of the global objective value and local iterative (projected) updates that use stochastic gradient estimates. We characterize how the closed-loop performance depends on the size of the network, the number of iterations, and the level of accuracy of consensus. Numerical simulations on a voltage control problem in a direct current power grid corroborate the theoretical findings.

We study the problem of preferential Bayesian optimization (BO), where we aim to optimize a black-box function with only preference feedback over a pair of candidate solutions. Inspired by the likelihood ratio idea, we construct a confidence set of the black-box function using only the preference feedback. An optimistic algorithm with an efficient computational method is then developed to solve the problem, which enjoys an information-theoretic bound on the total cumulative regret, a first-of-its-kind for preferential BO. This bound further allows us to design a scheme to report an estimated best solution, with a guaranteed convergence rate. Experimental results on sampled instances from Gaussian processes, standard test functions, and a thermal comfort optimization problem all show that our method stably achieves better or competitive performance as compared to the existing state-of-the-art heuristics, which, however, do not have theoretical guarantees on regret bounds or convergence., Comment: Accepted to ICML 2024

Sentiment analysis is rapidly advancing by utilizing various data modalities (e.g., text, image). However, most previous works relied on superficial information, neglecting the incorporation of contextual world knowledge (e.g., background information derived from but beyond the given image and text pairs) and thereby restricting their ability to achieve better multimodal sentiment analysis (MSA). In this paper, we proposed a plug-in framework named WisdoM, to leverage the contextual world knowledge induced from the large vision-language models (LVLMs) for enhanced MSA. WisdoM utilizes LVLMs to comprehensively analyze both images and corresponding texts, simultaneously generating pertinent context. To reduce the noise in the context, we also introduce a training-free contextual fusion mechanism. Experiments across diverse granularities of MSA tasks consistently demonstrate that our approach has substantial improvements (brings an average +1.96% F1 score among five advanced methods) over several state-of-the-art methods.

Online feedback optimization is a controller design paradigm for optimizing the steady-state behavior of a dynamical system. It employs an optimization algorithm as a dynamic feedback controller and utilizes real-time measurements to bypass knowing exact plant dynamics and disturbances. Different from existing centralized settings, we present a fully decentralized feedback optimization controller for networked systems to lift the communication burden and improve scalability. We approximate the overall input-output sensitivity matrix through its diagonal elements, which capture local model information. For the closed-loop behavior, we characterize the stability and bound the sub-optimality due to decentralization. We prove that the proposed decentralized controller yields solutions that correspond to the Nash equilibria of a non-cooperative game.

The pursuit of discovering new high-temperature superconductors that diverge from the copper-based paradigm1-3 carries profound implications for elucidating mechanisms behind superconductivity and may also enable new applications4-8. Here, our investigation reveals that application of pressure effectively suppresses the spin and charge order in trilayer nickelate La4Ni3O10-{\delta} single crystals, leading to the emergence of superconductivity with a maximum critical temperature (Tc) of around 30 K at 69.0 GPa. The DC susceptibility measurements confirm a substantial diamagnetic response below Tc, indicating the presence of bulk superconductivity with a volume fraction exceeding 80%. In the normal state, we observe a "strange metal" behavior, characterized by a linear temperature-dependent resistance extending up to 300 K. Furthermore, the layer-dependent superconductivity observed hints at a unique interlayer coupling mechanism specific to nickelates, setting them apart from cuprates in this regard. Our findings provide crucial insights into the fundamental mechanisms underpinning superconductivity, while also introducing a new material platform to explore the intricate interplay between the spin/charge order, flat band structures, interlayer coupling, strange metal behavior and high-temperature superconductivity., Comment: Supplementary information included

CNNs and Self attention have achieved great success in multimedia applications for dynamic association learning of self-attention and convolution in image restoration. However, CNNs have at least two shortcomings: 1) limited receptive field; 2) static weight of sliding window at inference, unable to cope with the content diversity.In view of the advantages and disadvantages of CNNs and Self attention, this paper proposes an association learning method to utilize the advantages and suppress their shortcomings, so as to achieve high-quality and efficient inpainting. We regard rain distribution reflects the degradation location and degree, in addition to the rain distribution prediction. Thus, we propose to refine background textures with the predicted degradation prior in an association learning manner. As a result, we accomplish image deraining by associating rain streak removal and background recovery, where an image deraining network and a background recovery network are designed for two subtasks. The key part of association learning is a novel multi-input attention module. It generates the degradation prior and produces the degradation mask according to the predicted rainy distribution. Benefited from the global correlation calculation of SA, MAM can extract the informative complementary components from the rainy input with the degradation mask, and then help accurate texture restoration. Meanwhile, SA tends to aggregate feature maps with self-attention importance, but convolution diversifies them to focus on the local textures. A hybrid fusion network involves one residual Transformer branch and one encoder-decoder branch. The former takes a few learnable tokens as input and stacks multi-head attention and feed-forward networks to encode global features of the image. The latter, conversely, leverages the multi-scale encoder-decoder to represent contexture knowledge., Comment: in Chinese language, Journal of Image and Graphics. arXiv admin note: substantial text overlap with arXiv:2207.10455

While most research into speech synthesis has focused on synthesizing high-quality speech for in-dataset speakers, an equally essential yet unsolved problem is synthesizing speech for unseen speakers who are out-of-dataset with limited reference data, i.e., speaker adaptive speech synthesis. Many studies have proposed zero-shot speaker adaptive text-to-speech and voice conversion approaches aimed at this task. However, most current approaches suffer from the degradation of naturalness and speaker similarity when synthesizing speech for unseen speakers (i.e., speakers not in the training dataset) due to the poor generalizability of the model in out-of-distribution data. To address this problem, we propose GZS-TV, a generalizable zero-shot speaker adaptive text-to-speech and voice conversion model. GZS-TV introduces disentangled representation learning for both speaker embedding extraction and timbre transformation to improve model generalization and leverages the representation learning capability of the variational autoencoder to enhance the speaker encoder. Our experiments demonstrate that GZS-TV reduces performance degradation on unseen speakers and outperforms all baseline models in multiple datasets., Comment: 5 pages, 3 figures. Accepted by Interspeech 2023, Oral

We use inelastic neutron scattering to study spin fluctuations in Sr$_{1.8}$La$_{0.2}$RuO$_4$, where Lanthanum doping triggers a Lifshitz transition by pushing the van Hove singularity in the $\gamma$ band to the Fermi energy. Strong spin fluctuations emerge at an incommensurate wave vector $\mathbf{Q}_{ic} = (0.3,0.3)$, corresponding to the nesting vector between $\alpha$ and $\beta$ Fermi sheets. The incommensurate antiferromagnetic fluctuations shift toward $(0.25,0.25)$ with increasing energy up to ${\sim}110$ meV. By contrast, scatterings near the ferromagnetic wave vectors $\mathbf{Q} = (1,0)$ and $(1,1)$ remain featureless at all energies. This contradicts the weak-coupling perspective that suggests a sharp enhancement of ferromagnetic susceptibility due to the divergence of density of states in the associated $\gamma$ band. Our findings imply that ferromagnetic fluctuations in Sr$_2$RuO$_4$ and related materials do not fit into the weak-coupling paradigm, but instead are quasi-local fluctuations induced by Hund's coupling. This imposes significant constraints for the pairing mechanism involving spin fluctuations., Comment: 7 pages, 4 figures

As an emerging approach, deep learning plays an increasingly influential role in channel modeling. Traditional ray tracing (RT) methods of channel modeling tend to be inefficient and expensive. In this paper, we present a super-resolution (SR) model for channel characteristics. Residual connection and attention mechanism are applied to this convolutional neural network (CNN) model. Experiments prove that the proposed model can reduce the noise interference generated in the SR process and solve the problem of low efficiency of RT. The mean absolute error of our channel SR model on the PL achieves the effect of 2.82 dB with scale factor 2, the same accuracy as RT took only 52\% of the time in theory. Compared with vision transformer (ViT), the proposed model also demonstrates less running time and computing cost in SR of channel characteristics.

Self-supervised facial representation has recently attracted increasing attention due to its ability to perform face understanding without relying on large-scale annotated datasets heavily. However, analytically, current contrastive-based self-supervised learning (SSL) still performs unsatisfactorily for learning facial representation. More specifically, existing contrastive learning (CL) tends to learn pose-invariant features that cannot depict the pose details of faces, compromising the learning performance. To conquer the above limitation of CL, we propose a novel Pose-disentangled Contrastive Learning (PCL) method for general self-supervised facial representation. Our PCL first devises a pose-disentangled decoder (PDD) with a delicately designed orthogonalizing regulation, which disentangles the pose-related features from the face-aware features; therefore, pose-related and other pose-unrelated facial information could be performed in individual subnetworks and do not affect each other's training. Furthermore, we introduce a pose-related contrastive learning scheme that learns pose-related information based on data augmentation of the same image, which would deliver more effective face-aware representation for various downstream tasks. We conducted linear evaluation on four challenging downstream facial understanding tasks, ie, facial expression recognition, face recognition, AU detection and head pose estimation. Experimental results demonstrate that our method significantly outperforms state-of-the-art SSL methods. Code is available at https://github.com/DreamMr/PCL}{https://github.com/DreamMr/PCL, Comment: Accepted by CVPR 2023

Strongly correlated materials often exhibit an electronic phase separation (EPS) phenomena whose domain pattern is random in nature. The ability to control the spatial arrangement of the electronic phases at microscopic scales is highly desirable for tailoring their macroscopic properties and/or designing novel electronic devices. Here we report the formation of EPS nanoscale network in a mono-atomically stacked LaMnO3/CaMnO3/PrMnO3 superlattice grown on SrTiO3 (STO) (001) substrate, which is known to have an antiferromagnetic (AFM) insulating ground state. The EPS nano-network is a consequence of an internal strain relaxation triggered by the structural domain formation of the underlying STO substrate at low temperatures. The same nanoscale network pattern can be reproduced upon temperature cycling allowing us to employ different local imaging techniques to directly compare the magnetic and transport state of a single EPS domain. Our results confirm the one-to-one correspondence between ferromagnetic (AFM) to metallic (insulating) state in manganite. It also represents a significant step in a paradigm shift from passively characterizing EPS in strongly correlated systems to actively engaging in its manipulation.

We propose an end-to-end lecture video generation system that can generate realistic and complete lecture videos directly from annotated slides, instructor's reference voice and instructor's reference portrait video. Our system is primarily composed of a speech synthesis module with few-shot speaker adaptation and an adversarial learning-based talking-head generation module. It is capable of not only reducing instructors' workload but also changing the language and accent which can help the students follow the lecture more easily and enable a wider dissemination of lecture contents. Our experimental results show that the proposed model outperforms other current approaches in terms of authenticity, naturalness and accuracy. Here is a video demonstration of how our system works, and the outcomes of the evaluation and comparison: https://youtu.be/cY6TYkI0cog., Comment: 4 pages, 4 figures, ICIP 2022

Dynamic facial expression recognition (FER) databases provide important data support for affective computing and applications. However, most FER databases are annotated with several basic mutually exclusive emotional categories and contain only one modality, e.g., videos. The monotonous labels and modality cannot accurately imitate human emotions and fulfill applications in the real world. In this paper, we propose MAFW, a large-scale multi-modal compound affective database with 10,045 video-audio clips in the wild. Each clip is annotated with a compound emotional category and a couple of sentences that describe the subjects' affective behaviors in the clip. For the compound emotion annotation, each clip is categorized into one or more of the 11 widely-used emotions, i.e., anger, disgust, fear, happiness, neutral, sadness, surprise, contempt, anxiety, helplessness, and disappointment. To ensure high quality of the labels, we filter out the unreliable annotations by an Expectation Maximization (EM) algorithm, and then obtain 11 single-label emotion categories and 32 multi-label emotion categories. To the best of our knowledge, MAFW is the first in-the-wild multi-modal database annotated with compound emotion annotations and emotion-related captions. Additionally, we also propose a novel Transformer-based expression snippet feature learning method to recognize the compound emotions leveraging the expression-change relations among different emotions and modalities. Extensive experiments on MAFW database show the advantages of the proposed method over other state-of-the-art methods for both uni- and multi-modal FER. Our MAFW database is publicly available from https://mafw-database.github.io/MAFW., Comment: This paper has been accepted by ACM MM'22

The structurally simplest high-temperature superconductor FeSe exhibits an intriguing superconducting nematic paramagnetic phase with unusual spin excitation spectra that are different from typical spin waves; thus, determining its effective magnetic exchange interactions is challenging. Here we report neutron scattering measurements of spin fluctuations of FeSe in the tetragonal paramagnetic phase. We show that the equal-time magnetic structure factor, $\mathcal{S}(\textbf{Q})$, can be effectively modeled using the self-consistent Gaussian approximation calculation with highly frustrated nearest-neighbor ($J_{1}$) and next-nearest-neighbor ($J_{2}$) exchange couplings, and very weak further neighbor exchange interaction. Our results elucidate the frustrated magnetism in FeSe, which provides a natural explanation for the highly tunable superconductivity and nematicity in FeSe and related materials.