IJCAI 2023 Papers

Refinement plays a critical role in language structure prediction, a process that deals with complex situations such as structural edge interdependencies. Since language structure prediction usually modeled as graph parsing, typical refinement methods involve taking an initial parsing graph as input and refining it using language input and other relevant information. Intuitively, a refinement component, i.e., refiner, should be lightweight and efficient, as it is only responsible for correcting faults in the initial graph. However, current refiners add a significant burden to the parsing process due to their reliance on time-consuming encoding-decoding procedure on the language input and graph. To make the refiner more practical for real-world applications, this paper proposes a lightweight but effective iterative refinement framework, iRe^2f, based on iterative retrospecting and reasoning without involving the re-encoding process on the graph. iRe^2f iteratively refine the parsing graph based on interaction between graph and sequence and efficiently learns the shortcut to update the sequence and graph representations in each iteration. The shortcut is calculated based on the graph representation in the latest iteration. iRe^2f reduces the number of refinement parameters by 90% compared to the previous smallest refiner. Experiments on a variety of language structure prediction tasks show that iRe^2f performs comparably or better than current state-of-the-art refiners, with a significant increase in efficiency.

Hyperparameter optimization (HPO) is crucial for strong performance of deep learning algorithms and real-world applications often impose some constraints, such as memory usage, or latency on top of the performance requirement. In this work, we propose constrained TPE (c-TPE), an extension of the widely-used versatile Bayesian optimization method, tree-structured Parzen estimator (TPE), to handle these constraints. Our proposed extension goes beyond a simple combination of an existing acquisition function and the original TPE, and instead includes modifications that address issues that cause poor performance. We thoroughly analyze these modifications both empirically and theoretically, providing insights into how they effectively overcome these challenges. In the experiments, we demonstrate that c-TPE exhibits the best average rank performance among existing methods with statistical significance on 81 expensive HPO with inequality constraints. Due to the lack of baselines, we only discuss the applicability of our method to hard-constrained optimization in Appendix D. See https://arxiv.org/abs/2211.14411 for the latest version with Appendix.

We present XFormer, a novel human mesh and motion capture method that achieves real-time performance on consumer CPUs given only monocular images as input. The proposed network architecture contains two branches: a keypoint branch that estimates 3D human mesh vertices given 2D keypoints, and an image branch that makes prediction directly from the RGB image features. At the core of our method is a cross-modal transformer block that allows information flow across these two branches by modeling the attention between 2D keypoint coordinates and image spatial features. Our architecture is smartly designed, which enables us to train on various types of datasets including images with 2D/3D annotations, images with 3D pseudo labels, and motion capture datasets that do not have associated images. This effectively improves the accuracy and generalization ability of our system. Built on a lightweight backbone (MobileNetV3), our method runs blazing fast (over 30fps on a single CPU core) and still yields competitive accuracy. Furthermore, with a HRNet backbone, XFormer delivers state-of-the-art performance on Huamn3.6 and 3DPW datasets.

The top-down and bottom-up methods are two mainstreams of referring segmentation, while both methods have their own intrinsic weaknesses. Top-down methods are chiefly disturbed by Polar Negative (PN) errors owing to the lack of fine-grained cross-modal alignment. Bottom-up methods are mainly perturbed by Inferior Positive (IP) errors due to the lack of prior object information. Nevertheless, we discover that two types of methods are highly complementary for restraining respective weaknesses but the direct average combination leads to harmful interference. In this context, we build Win-win Cooperation (WiCo) to exploit complementary nature of two types of methods on both interaction and integration aspects for achieving a win-win improvement. For the interaction aspect, Complementary Feature Interaction (CFI) introduces prior object information to bottom-up branch and provides fine-grained information to top-down branch for complementary feature enhancement. For the integration aspect, Gaussian Scoring Integration (GSI) models the gaussian performance distributions of two branches and weighted integrates results by sampling confident scores from the distributions. With our WiCo, several prominent bottom-up and top-down combinations achieve remarkable improvements on three common datasets with reasonable extra costs, which justifies effectiveness and generality of our method.

We study a rumor spreading model where individuals are connected via a network structure. Initially, only a small subset of the individuals are spreading a rumor. Each individual who is connected to a spreader, starts spreading the rumor with some probability as a function of their trust in the spreader, quantified by the Jaccard similarity index. Furthermore, the probability that a spreader diffuses the rumor decreases over time until they fully lose their interest and stop spreading. We focus on determining the graph parameters which govern the magnitude and pace that the rumor spreads in this model. We prove that for the rumor to spread to a sizable fraction of the individuals, the network needs to enjoy ``strong'' expansion properties and most nodes should be in ``well-connected'' communities. Both of these characteristics are, arguably, present in real-world social networks up to a certain degree, shedding light on the driving force behind the extremely fast spread of rumors in social networks. Furthermore, we formulate a large range of countermeasures to cease the spread of a rumor. We introduce four fundamental criteria which a countermeasure ideally should possess. We evaluate all the proposed countermeasures by conducting experiments on real-world social networks such as Facebook and Twitter. We conclude that our novel decentralized countermeasures (which are executed by the individuals) generally outperform the previously studied centralized ones (which need to be imposed by a third entity such as the government).

We present a review that unifies decision-support methods for exploring the solutions produced by multi-objective optimization (MOO) algorithms. As MOO is applied to solve diverse problems, approaches for analyzing the trade-offs offered by these algorithms are scattered across fields. We provide an overview of the current advances on this topic, including methods for visualization, mining the solution set, and uncertainty exploration as well as emerging research directions, including interactivity, explainability, and support on ethical aspects. We synthesize these methods drawing from different fields of research to enable building a unified approach, independent of the application. Our goals are to reduce the entry barrier for researchers and practitioners on using MOO algorithms and to provide novel research directions.

The sRGB white balance methods aim to correct the nonlinear color cast of sRGB images without accessing raw values. Although existing methods have achieved increasingly better results, their generalization to sRGB images from multiple cameras is still under explored. In this paper, we propose the network named WBFlow that not only performs superior white balance for sRGB images but also generalizes well to multiple cameras. Specifically, we take advantage of neural flow to ensure the reversibility of WBFlow, which enables lossless rendering of color cast sRGB images back to pseudo raw features for linear white balancing and thus achieves superior performance. Furthermore, inspired by camera transformation approaches, we have designed a camera transformation (CT) in pseudo raw feature space to generalize WBFlow for different cameras via few shot learning. By utilizing a few sRGB images from an untrained camera, our WBFlow can perform well on this camera by learning the camera specific parameters of CT. Extensive experiments show that WBFlow achieves superior camera generalization and accuracy on three public datasets as well as our rendered multiple camera sRGB dataset. Our code is available at https://github.com/ChunxiaoLe/WBFlow.

Adversarial example is a rising tool for voice privacy protection. By adding imperceptible noise to public audio, it prevents tampers from using zero-shot Voice Conversion (VC) to synthesize high quality speech with target speaker identity. However, many existing studies ignore the human perception characteristics of audio data, and it is challenging to generate strong and imperceptible adversarial audio. In this paper, we propose the Voice Guard defense method, which uses a novel method to advance the adversarial perturbation to the time domain to avoid the loss caused by cross-domain conversion. And the psychoacoustic model is introduced into the defense of VC for the first time, which greatly improves the disruption ability and concealment of adversarial audio. We also standardize the evaluation metrics of adversarial audio for the first time, combining multi-dimensional metrics to define the criteria for defense. We evaluate Voice Guard on several state-of-the-art zero-shot VC models. The experimental results show that our method can ensure the perceptual quality of adversarial audio while having a strong defense capability, and is far superior to previous works in terms of disruption ability and concealment.

Vision-language navigation (VLN) requires an agent to perceive visual observation in a house scene and navigate step-by-step following natural language instruction. Due to the high cost of data annotation and data collection, current VLN datasets provide limited instruction-trajectory data samples. Learning vision-language alignment for VLN from limited data is challenging since visual observation and language instruction are both complex and diverse. Previous works only generate augmented data based on original scenes while failing to generate data samples from unseen scenes, which limits the generalization ability of the navigation agent. In this paper, we introduce the Knowledge-driven Environmental Dreamer (KED), a method that leverages the knowledge of the embodied environment and generates unseen scenes for a navigation agent to learn. Generating an unseen environment with texture consistency and structure consistency is challenging. To address this problem, we incorporate three knowledge-driven regularization objectives into the KED and adopt a reweighting mechanism for self-adaptive optimization. Our KED method is able to generate unseen embodied environments without extra annotations. We use KED to successfully generate 270 houses and 500K instruction-trajectory pairs. The navigation agent with the KED method outperforms the state-of-the-art methods on various VLN benchmarks, such as R2R, R4R, and RxR. Both qualitative and quantitative experiments prove that our proposed KED method is able to high-quality augmentation data with texture consistency and structure consistency.

Graph Neural Networks (GNNs) is a family of promising tools for graph semi-supervised learning. However, in training, most existing GNNs rely heavily on a large amount of labeled data, which is rare in real-world scenarios. Unlabeled data with useful information are usually under-exploited, which limits the representation power of GNNs. To handle these problems, we propose Virtual Overbridge Linking (Violin), a generic framework to enhance the learning capacity of common GNNs. By learning to add virtual overbridges between two nodes that are estimated to be semantic-consistent, labeled and unlabeled data can be correlated. Supervised information can be well utilized in training while simultaneously inducing the model to learn from unlabeled data. Discriminative relation patterns extracted from unlabeled nodes can also be shared with other nodes even if they are remote from each other. Motivated by recent advances in data augmentations, we additionally integrate Violin with the consistency regularized training. Such a scheme yields node representations with better robustness, which significantly enhances a GNN. Violin can be readily extended to a wide range of GNNs without introducing additional learnable parameters. Extensive experiments on six datasets demonstrate that our method is effective and robust under low-label rate scenarios, where Violin can boost some GNNs' performance by over 10% on node classifications.

To free human from laborious video production, this paper proposes the building of VideoMaster, a multimodal system equipped with four capabilities: highlight extraction, video describing, video dubbing and video editing. It extracts interesting episodes from long game videos, generates subtitles for each episode, reads the subtitles through synthesized speech, and finally re-creates a better short video through video editing. Notably, VideoMaster takes a combination of deep learning and traditional computer vision techniques to extract highlights with fine-to-coarse labels, utilizes a novel framework named PCSG-v (probabilistic context sensitive grammar for video) for video description generation, and imitates a target speaker's voice to read the description. To the best of our knowledge, VideoMaster is the first multimedia system that can automatically produce product-level micro-videos without heavy human annotation.

In this paper, we introduce semi-supervised video object segmentation (VOS) to panoptic wild scenes and present a large-scale benchmark as well as a baseline method for it. Previous benchmarks for VOS with sparse annotations are not sufficient to train or evaluate a model that needs to process all possible objects in real-world scenarios. Our new benchmark (VIPOSeg) contains exhaustive object annotations and covers various real-world object categories which are carefully divided into subsets of thing/stuff and seen/unseen classes for comprehensive evaluation. Considering the challenges in panoptic VOS, we propose a strong baseline method named panoptic object association with transformers (PAOT), which associates multiple objects by panoptic identification in a pyramid architecture on multiple scales. Experimental results show that VIPOSeg can not only boost the performance of VOS models by panoptic training but also evaluate them comprehensively in panoptic scenes. Previous methods for classic VOS still need to improve in performance and efficiency when dealing with panoptic scenes, while our PAOT achieves SOTA performance with good efficiency on VIPOSeg and previous VOS benchmarks. PAOT also ranks 1st in the VOT2022 challenge. Our dataset and code are available at https://github.com/yoxu515/VIPOSeg-Benchmark.

Video Frame Interpolation (VFI) aims to synthesize non-existent intermediate frames between existent frames. Flow-based VFI algorithms estimate intermediate motion fields to warp the existent frames. Real-world motions' complexity and the reference frame's absence make motion estimation challenging. Many state-of-the-art approaches explicitly model the correlations between two neighboring frames for more accurate motion estimation. In common approaches, the receptive field of correlation modeling at higher resolution depends on the motion fields estimated beforehand. Such receptive field dependency makes common motion estimation approaches poor at coping with small and fast-moving objects. To better model correlations and to produce more accurate motion fields, we propose the Densely Queried Bilateral Correlation (DQBC) that gets rid of the receptive field dependency problem and thus is more friendly to small and fast-moving objects. The motion fields generated with the help of DQBC are further refined and up-sampled with context features. After the motion fields are fixed, a CNN-based SynthNet synthesizes the final interpolated frame. Experiments show that our approach enjoys higher accuracy and less inference time than the state-of-the-art. Source code is available at https://github.com/kinoud/DQBC.

Diffusion models have emerged as a powerful paradigm in video synthesis tasks including prediction, generation, and interpolation. Due to the limitation of the computational budget, existing methods usually implement conditional diffusion models with an autoregressive inference pipeline, in which the future fragment is predicted based on the distribution of adjacent past frames. However, only the conditions from a few previous frames can't capture the global temporal coherence, leading to inconsistent or even outrageous results in long-term video prediction. In this paper, we propose a Local-Global Context guided Video Diffusion model (LGC-VD) to capture multi-perception conditions for producing high-quality videos in both conditional/unconditional settings. In LGC-VD, the UNet is implemented with stacked residual blocks with self-attention units, avoiding the undesirable computational cost in 3D Conv. We construct a local-global context guidance strategy to capture the multi-perceptual embedding of the past fragment to boost the consistency of future prediction. Furthermore, we propose a two-stage training strategy to alleviate the effect of noisy frames for more stable predictions. Our experiments demonstrate that the proposed method achieves favorable performance on video prediction, interpolation, and unconditional video generation. We release code at https://github.com/exisas/LGC-VD.

Multi-camera instance association, which identifies identical objects among multiple objects in multi-view images, is challenging due to several harsh constraints. To tackle this problem, most studies have employed CNNs as feature extractors but often fail under such harsh constraints. Inspired by Vision Transformer (ViT), we first develop a pure ViT-based framework for robust feature extraction through self-attention and residual connection. We then propose two novel methods to achieve robust feature learning. First, we introduce learnable pseudo 3D position embeddings (P3DEs) that represent the 3D location of an object in the world coordinate system, which is independent of the harsh constraints. To generate P3DEs, we encode the camera ID and the object's 2D position in the image using embedding tables. We then build a framework that trains P3DEs to represent an object's 3D position in a weakly supervised manner. Second, we also utilize joint patch generation (JPG). During patch generation, JPG considers an object and its surroundings as a single input patch to reinforce the relationship information between two features. Ultimately, experimental results demonstrate that both ViT-P3DE and ViT-P3DE with JPG achieve state-of-the-art performance and significantly outperform existing works, especially when dealing with extremely harsh constraints.

Despite the popularity of Vision Transformers (ViTs) and eXplainable AI (XAI), only a few explanation methods have been designed specially for ViTs thus far. They mostly use attention weights of the [CLS] token on patch embeddings and often produce unsatisfactory saliency maps. This paper proposes a novel method for explaining ViTs called ViT-CX. It is based on patch embeddings, rather than attentions paid to them, and their causal impacts on the model output. Other characteristics of ViTs such as causal overdetermination are considered in the design of ViT-CX. The empirical results show that ViT-CX produces more meaningful saliency maps and does a better job revealing all important evidence for the predictions than previous methods. The explanation generated by ViT-CX also shows significantly better faithfulness to the model. The codes and appendix are available at https://github.com/vaynexie/CausalX-ViT.

Due to the insufficiency of electronic health records (EHR) data utilized in practical diagnosis prediction scenarios, most works are devoted to learning powerful patient representations either from structured EHR data (e.g., temporal medical events, lab test results, etc.) or unstructured data (e.g., clinical notes, etc.). However, synthesizing rich information from both of them still needs to be explored. Firstly, the heterogeneous semantic biases across them heavily hinder the synthesis of representation spaces, which is critical for diagnosis prediction. Secondly, the intermingled quality of partial clinical notes leads to inadequate representations of to-be-predicted patients. Thirdly, typical attention mechanisms mainly focus on aggregating information from similar patients, ignoring important auxiliary information from others. To tackle these challenges, we propose a novel visit sequences-clinical notes joint learning approach, dubbed VecoCare. It performs a Gromov-Wasserstein Distance (GWD)-based contrastive learning task and an adaptive masked language model task in a sequential pre-training manner to reduce heterogeneous semantic biases. After pre-training, VecoCare further aggregates information from both similar and dissimilar patients through a dual-channel retrieval mechanism. We conduct diagnosis prediction experiments on two real-world datasets, which indicates that VecoCare outperforms state-of-the-art approaches. Moreover, the findings discovered by VecoCare are consistent with the medical researches.

Unlike conventional zero-shot learning (CZSL) which only focuses on the recognition of unseen classes by using the classifier trained on seen classes and semantic embeddings, generalized zero-shot learning (GZSL) aims at recognizing both the seen and unseen classes, so it is more challenging due to the extreme training imbalance. Recently, some feature generation methods introduce metric learning to enhance the discriminability of visual features. Although these methods achieve good results, they focus only on metric learning in the visual feature space to enhance features and ignore the association between the feature space and the semantic space. Since the GZSL method uses semantics as prior knowledge to migrate visual knowledge to unseen classes, the consistency between visual space and semantic space is critical. To this end, we propose relational metric learning which can relate the metrics in the two spaces and make the distribution of the two spaces more consistent. Based on the generation method and relational metric learning, we proposed a novel GZSL method, termed VS-Boost, which can effectively boost the association between vision and semantics. The experimental results demonstrate that our method is effective and achieves significant gains on five benchmark datasets compared with the state-of-the-art methods.

Neural Radiance Fields (NeRF) has shown great success in novel view synthesis due to its state-of-the-art quality and flexibility. However, NeRF requires dense input views (tens to hundreds) and a long training time (hours to days) for a single scene to generate high-fidelity images. Although using the voxel grids to represent the radiance field can significantly accelerate the optimization process, we observe that for sparse inputs, the voxel grids are more prone to overfitting to the training views and will have holes and floaters, which leads to artifacts. In this paper, we propose VGOS, an approach for fast (3-5 minutes) radiance field reconstruction from sparse inputs (3-10 views) to address these issues. To improve the performance of voxel-based radiance field in sparse input scenarios, we propose two methods: (a) We introduce an incremental voxel training strategy, which prevents overfitting by suppressing the optimization of peripheral voxels in the early stage of reconstruction. (b) We use several regularization techniques to smooth the voxels, which avoids degenerate solutions. Experiments demonstrate that VGOS achieves state-of-the-art performance for sparse inputs with super-fast convergence. Code will be available at https://github.com/SJoJoK/VGOS.