Modeling how humans interact with objects is crucial for AI to effectively assist or mimic human behaviors. Existing studies for learning such ability primarily focus on static human-object interaction (HOI) patterns, such as contact and spatial relationships, while dynamic HOI patterns, capturing the movement of humans and objects over time, remain relatively underexplored. In this paper, we present a novel framework for learning Dynamic Affordance across various target object categories. To address the scarcity of 4D HOI datasets, our method learns the 3D dynamic affordance from synthetically generated 4D HOI samples. Specifically, we propose a pipeline that first generates 2D HOI videos from a given 3D target object using a pre-trained video diffusion model, then lifts them into 3D to generate 4D HOI samples. Leveraging these synthesized 4D HOI samples, we train DAViD, our generative 4D human-object interaction model, which is composed of two key components: (1) a human motion diffusion model (MDM) with Low-Rank Adaptation (LoRA) module to fine-tune a pre-trained MDM to learn the HOI motion concepts from limited HOI motion samples, (2) a motion diffusion model for 4D object poses conditioned by produced human interaction motions. Interestingly, DAViD can integrate newly learned HOI motion concepts with pre-trained human motions to create novel HOI motions, even for multiple HOI motion concepts, demonstrating the advantage of our pipeline with LoRA in integrating dynamic HOI concepts. Through extensive experiments, we demonstrate that DAViD outperforms baselines in synthesizing HOI motion.

Chanjun Park, Hyeonwoo Kim, Dahyun Kim, SeongHwan Cho, Sanghoon Kim, Sukyung Lee, Yungi Kim, Hwalsuk Lee. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

We propose Text2Scene, a method to automatically create realistic textures for virtual scenes composed of multiple objects. Guided by a reference image and text descriptions, our pipeline adds detailed texture on labeled 3D geometries in the room such that the generated colors respect the hierarchical structure or semantic parts that are often composed of similar materials. Instead of applying flat stylization on the entire scene at a single step, we obtain weak semantic cues from geometric segmentation, which are further clarified by assigning initial colors to segmented parts. Then we add texture details for individual objects such that their projections on image space exhibit feature embedding aligned with the embedding of the input. The decomposition makes the entire pipeline tractable to a moderate amount of computation resources and memory. As our framework utilizes the existing resources of image and text embedding, it does not require dedicated datasets with high-quality textures designed by skillful artists. To the best of our knowledge, it is the first practical and scalable approach that can create detailed and realistic textures of the desired style that maintain structural context for scenes with multiple objects.

Due to the spread of various information and communication technologies, a huge amount of images are produced and shared for diverse purposes. Several de-identification techniques for photos, such as pixelation, blur, and mask, are routinely used in light of recent worries about the growing number of privacy leakages. However, due to the low image quality and loss of many facial features, these de-identified images are not suitable for use in applications such as training models that require a lot of high-quality data. Therefore, in this paper, we propose a new face de-identification method focusing only on facial regions essential for personal identification. By generating facial landmarks differently from the original person using masking and generative adversarial networks-based inpainting, our method can perform de-identification efficiently. To demonstrate the performance of our proposed scheme, we conducted quantitative and qualitative evaluations using an open dataset. We show that our proposed scheme outperforms other de-identification methods.

Photo cartoonization aims to convert photos of real-world scenes into cartoon-style images. Recently, generative adversarial network (GAN)-based methods for photo cartoonization have been proposed to generate pleasable cartoonized images. However, as these methods can transfer only learned cartoon styles to photos, they are limited in general-purpose applications where unlearned styles are often required. To address this limitation, an arbitrary style transfer (AST) method that transfers arbitrary artistic style into content images can be used. However, conventional AST methods do not perform satisfactorily in cartoonization for two reasons. First, they cannot capture the unique characteristics of cartoons that differ from common artistic styles. Second, they suffer from content leaks in which the semantic structure of the content is distorted. In this paper, to solve these problems, we propose a novel arbitrary-style photo cartoonization method, Cartoon-Flow. More specifically, we construct a new hybrid GAN with an invertible neural flow generator to effectively preserve content information. In addition, we introduce two new losses for cartoonization: (1) edge-promoting smooth loss to learn the unique characteristics of cartoons with smooth surfaces and clear edges, and (2) line loss to mimic the line drawing of cartoons. Extensive experiments demonstrate that the proposed method outperforms previous methods both quantitatively and qualitatively.

In this paper, we propose a proactive security method that estimates distributed denial of service (DDoS) attack volume in order to overcome the limitation of the response time of reactive security systems based on intrusion detection. To that end, we define and discuss network intrusion forecasting and intrusion factors in comparison to intrusion detection. Intrusion factors for a DDoS attack are also analyzed and collected from the Honeynet system. Based on the data from the Honeynet system, we conduct correlation and regression analysis, both statistical approaches, to predict the potential DDoS attack volume for the network security of our university. By combining network intrusion detection with intrusion forecasting, network operators can take active countermeasures based on the forecasting results and strengthen the network security.

We report performance measurements and analyses of a variety of wireless networks to arrive at an appropriate network structure configuration for smooth multimedia streaming service with a variety of smart devices in a single wireless network environment. Unlike the usual infrastructure network, configurations such as the IEEE 802.11 ad hoc network and the Wi-Fi direct network use direct connections between devices without going through a wireless AP. Therefore, these configurations prevent concentrating traffic at the wireless AP. We generated three types of wireless network performance measurements and arrived at a suitable network for multimedia streaming service using a variety of smart devices. The wireless network was configured in a structure for efficient multimedia streaming service in a single wireless LAN environment, and subsequently, the network performance was measured.

The method of inferring firewall policy, using Active Probing repeats the process of transmitting TCP/UDP/ICMP packets and receiving ICMP response packets. However, if ICMP response packets cannot be received normally, the accuracy of inferring the firewall policy decreases, and it is necessary to verify the feasibility in real conditions. In this paper, we collect Autonomous System (AS) information to investigate the tolerance of ICMP intended for all AS across the world in addition to DNS server information, which is operational within AS. We confirm whether ICMP response packets are received or not by transmitting probing packets to the DNS server. Finally, we propose the AS information that received ICMP packets as the result of the test.

We propose a framework which infers the policy of firewall deployed in the Internet access point and computer system. The proposed methodology shows how to infer a firewall policy from restricted probing packets, using consecutive characteristics of the IP address and TCP/UDP port number. We also show the experimental results and the performance of the proposed method.