Weakly supervised video grounding aims to localize temporal boundaries relevant to a given query without explicit ground-truth temporal boundaries. While existing methods primarily use Gaussian-based proposals, they overlook the importance of (1) boundary prediction and (2) top-1 prediction selection during inference. In their boundary prediction, boundaries are simply set at half a standard deviation away from a Gaussian mean on both sides, which may not accurately capture the optimal boundaries. In the top-1 prediction process, these existing methods rely heavily on intersections with other proposals, without considering the varying quality of each proposal. To address these issues, we explore various inference strategies by introducing (1) novel boundary prediction methods to capture diverse boundaries from multiple Gaussians and (2) new selection methods that take proposal quality into account. Extensive experiments on the ActivityNet Captions and Charades-STA datasets validate the effectiveness of our inference strategies, demonstrating performance improvements without requiring additional training.

Recently, open-source large Vision-Language Models (VLMs) have progressed toward achieving performance comparable to closed-source VLMs. However, open-source VLMs struggle to recognize unfamiliar texts depicted in the images, where the texts are written in underrepresented languages like Korean. This illiteracy problem is primarily due to insufficient training data for the underrepresented languages. To address this problem, we propose a novel augmentation scheme that generates large-scale image data for the underrepresented languages with minimal manual annotations. Our scheme synthetically combines a text image depicting words or sentences with a template image containing textual contexts, such as a receipt, a sign, a book, and a product label. Specifically, the text image is cut and pasted into a patch of the template image to generate a synthetic image, which is labeled with the corresponding texts in the text image. Therefore, fine-tuning a VLM with our synthetic data can enhance its ability to generalize to real-world text recognition tasks. Experimental results demonstrate the effectiveness of our scheme, showing a significant performance improvement in text recognition.

Node-level out-of-distribution (OOD) detection on graphs has received significant attention from the machine learning community. However, previous approaches are evaluated using unrealistic benchmarks that consider only randomly selected OOD nodes, failing to reflect the interactions among nodes. In this paper, we introduce a new challenging task to model the interactions of OOD nodes in a graph, termed spreading OOD detection, where a newly emerged OOD node spreads its property to neighboring nodes. We curate realistic benchmarks by employing the epidemic spreading models that simulate the spreading of OOD nodes on the graph. We also showcase a ``Spreading COVID-19" dataset to demonstrate the applicability of spreading OOD detection in real-world scenarios. Furthermore, to effectively detect spreading OOD samples under the proposed benchmark setup, we present a new approach called energy distribution-based detector (EDBD), which includes a novel energy-aggregation scheme. EDBD is designed to mitigate undesired mixing of OOD scores between in-distribution (ID) and OOD nodes. Our extensive experimental results demonstrate the superiority of our approach over state-of-the-art methods in both spreading OOD detection and conventional node-level OOD detection tasks across seven benchmark datasets. The source code is available at https://github.com/daehoum1/edbd.

In this paper, we address learning tasks on graphs with missing features, enhancing the applicability of graph neural networks to real-world graph-structured data. We identify a critical limitation of existing imputation methods based on feature propagation: they produce channels with nearly identical values within each channel, and these low-variance channels contribute very little to performance in graph learning tasks. To overcome this issue, we introduce synthetic features that target the root cause of low-variance channel production, thereby increasing variance in these channels. By preventing propagation-based imputation methods from generating meaningless feature values shared across all nodes, our synthetic feature propagation scheme mitigates significant performance degradation, even under extreme missing rates. Extensive experiments demonstrate the effectiveness of our approach across various graph learning tasks with missing features, ranging from low to extremely high missing rates. Additionally, we provide both empirical evidence and theoretical proof to validate the low-variance problem. The source code is available at https://github.com/daehoum1/fisf.

In the weakly supervised temporal video grounding study, previous methods use predetermined single Gaussian proposals which lack the ability to express diverse events described by the sentence query. To enhance the expression ability of a proposal, we propose a Gaussian mixture proposal (GMP) that can depict arbitrary shapes by learning importance, centroid, and range of every Gaussian in the mixture. In learning GMP, each Gaussian is not trained in a feature space but is implemented over a temporal location. Thus the conventional feature-based learning for Gaussian mixture model is not valid for our case. In our special setting, to learn moderately coupled Gaussian mixture capturing diverse events, we newly propose a pull-push learning scheme using pulling and pushing losses, each of which plays an opposite role to the other. The effects of components in our scheme are verified in-depth with extensive ablation studies and the overall scheme achieves state-of-the-art performance. Our code is available at https://github.com/sunoh-kim/pps.

In this paper, to effectively capture non-sequential dependencies among semantic words for temporal video grounding, we propose a novel framework called Semantic Word-Aware Graph Network (SWAG-Net), which adopts graph-guided semantic word embedding in an end-to-end manner. Specifically, we define semantic word features as node features of semantic word-aware graphs and word-to-word correlations as three edge types (i.e., intrinsic, extrinsic, and relative edges) for diverse graph structures. We then apply Semantic Word-aware Graph Convolutional Networks (SW-GCNs) to the graphs for semantic word embedding. For modality fusion and context modeling, the embedded features and video segment features are merged into bi-modal features, and the bi-modal features are aggregated by incorporating local and global contextual information. Leveraging the aggregated features, the proposed method effectively finds a temporal boundary semantically corresponding to a sentence query in an untrimmed video. We verify that our SWAG-Net outperforms state-of-the-art methods on Charades-STA and ActivityNet Captions datasets.

The key to successful grounding for video surveillance is to understand a semantic phrase corresponding to important actors and objects. Conventional methods ignore comprehensive contexts for the phrase or require heavy computation for multiple phrases. To understand comprehensive contexts with only one semantic phrase, we propose Position-aware Location Regression Network (PLRN) which exploits position-aware features of a query and a video. Specifically, PLRN first encodes both the video and query using positional information of words and video segments. Then, a semantic phrase feature is extracted from an encoded query with attention. The semantic phrase feature and encoded video are merged and made into a context-aware feature by reflecting local and global contexts. Finally, PLRN predicts start, end, center, and width values of a grounding boundary. Our experiments show that PLRN achieves competitive performance over existing methods with less computation time and memory.

In visual surveillance systems, it is necessary to recognize the behavior of people handling objects such as a phone, a cup, or a plastic bag. In this paper, to address this problem, we propose a new framework for recognizing object-related human actions by graph convolutional networks using human and object poses. In this framework, we construct skeletal graphs of reliable human poses by selectively sampling the informative frames in a video, which include human joints with high confidence scores obtained in pose estimation. The skeletal graphs generated from the sampled frames represent human poses related to the object position in both the spatial and temporal domains, and these graphs are used as inputs to the graph convolutional networks. Through experiments over an open benchmark and our own data sets, we verify the validity of our framework in that our method outperforms the state-of-the-art method for skeleton-based action recognition.