Abstract: Generative models have advanced significantly, synthesizing photorealistic images, videos, and text. Building on this progress, our work explores the potential of diffusion models to bridge generative and discriminative learning, uncovering new pathways for leveraging their strengths in visual perception tasks. In the first part, we propose Diff-2-in-1, a unified framework for multi-modal data generation [...]

Abstract: Uncrewed Aerial Vehicles (UAVs) have attracted the interest of researchers, industry, and the general public in many applications. Noticing that high-altitude tasks sometimes require active interaction with the environment, there have been more and more works focusing on aerial manipulation recently. Each of them has demonstrated the ability to use a specific aerial manipulator [...]

Abstract: Autonomous robot exploration is widely applied in areas such as search and rescue, environmental monitoring, and structural inspection. Multi-robot exploration has garnered significant attention in the robotics research community, as it enables faster task completion and greater coverage than a single robot can achieve. However, it presents unique challenges: behavior coordination is complex, communication [...]

Abstract: In an age of autonomous driving and robotics, we are increasingly engaging with robots that deploy autonomous assistance. Cognitive science and human-computer interaction literature tells us that, when we apply autonomy in assistive settings, we are often augmenting the user's sense of agency over the system. Sense of agency is a phenomenon from cognitive [...]

Abstract: Medical emergencies demand rapid and accurate interventions to save lives. Severe injuries often require surgical care within the first 60 minutes when timely action significantly improves survival rates. However, limited resources, remote locations, and unpredictable conditions often obstruct access to advanced medical care during this critical period. This thesis focuses on developing a medical [...]

Abstract: Human-centered computer vision technology relies heavily on large, diverse datasets, but collecting data from human subjects is time-consuming, labor-intensive, and raises privacy concerns. To address these challenges, researchers are increasingly using synthetic data to augment real-world datasets. This thesis explores efficient methods for generating and utilizing synthetic data to train human-based computer vision models. [...]

Abstract: Visual imitation learning has shown promising results in robotic manipulation in recent years. However, its generalization to unseen objects is often limited by the size and diversity of training data. Although more large-scale robotic datasets are available, they remain significantly smaller than image and text datasets. Additionally, scaling these datasets is time-consuming and labor-intensive, [...]