Christoph Gebhardt

Ari Blau, Christoph Gebhardt, Andres Bendesky et al.

Multi-animal pose estimation is essential for studying animals' social behaviors in neuroscience and neuroethology. Advanced approaches have been proposed to support multi-animal estimation and achieve state-of-the-art performance. However, these models rarely exploit unlabeled data during training even though real world applications have exponentially more unlabeled frames than labeled frames. Manually adding dense annotations for a large number of images or videos is costly and labor-intensive, especially for multiple instances. Given these deficiencies, we propose a novel semi-supervised architecture for multi-animal pose estimation, leveraging the abundant structures pervasive in unlabeled frames in behavior videos to enhance training, which is critical for sparsely-labeled problems. The resulting algorithm will provide superior multi-animal pose estimation results on three animal experiments compared to the state-of-the-art baseline and exhibits more predictive power in sparsely-labeled data regimes.

Zhipeng Li, Christoph Gebhardt, Yi-Chi Liao et al.

We present AutoOptimization, a novel multi-objective optimization framework for adapting user interfaces. From a user's verbal preferences for changing a UI, our framework guides a prioritization-based Pareto frontier search over candidate layouts. It selects suitable objective functions for UI placement while simultaneously parameterizing them according to the user's instructions to define the optimization problem. A solver then generates a series of optimal UI layouts, which our framework validates against the user's instructions to adapt the UI with the final solution. Our approach thus overcomes the previous need for manual inspection of layouts and the use of population averages for objective parameters. We integrate multiple agents sequentially within our framework, enabling the system to leverage their reasoning capabilities to interpret user preferences, configure the optimization problem, and validate optimization outcomes.

Christoph Gebhardt, Robin Willardt, Seyedmorteza Sadat et al.

Emotions are known to mediate the relationship between users' content consumption and their online engagement, with heightened emotional intensity leading to increased engagement. Building on this insight, we propose three regressor-guided image editing approaches aimed at diminishing the emotional impact of images. These include (i) a parameter optimization approach based on global image transformations known to influence emotions, (ii) an optimization approach targeting the style latent space of a generative adversarial network, and (iii) a diffusion-based approach employing classifier guidance and classifier-free guidance. Our findings demonstrate that approaches can effectively alter the emotional properties of images while maintaining high visual quality. Optimization-based methods primarily adjust low-level properties like color hues and brightness, whereas the diffusion-based approach introduces semantic changes, such as altering appearance or facial expressions. Notably, results from a behavioral study reveal that only the diffusion-based approach successfully elicits changes in viewers' emotional responses while preserving high perceived image quality. In future work, we will investigate the impact of these image adaptations on internet user behavior.

Mert Albaba, Sammy Christen, Thomas Langarek et al.

Acquiring complex behaviors is essential for artificially intelligent agents, yet learning these behaviors in high-dimensional settings poses a significant challenge due to the vast search space. Traditional reinforcement learning (RL) requires extensive manual effort for reward function engineering. Inverse reinforcement learning (IRL) uncovers reward functions from expert demonstrations but relies on an iterative process that is often computationally expensive. Imitation learning (IL) provides a more efficient alternative by directly comparing an agent's actions to expert demonstrations; however, in high-dimensional environments, such direct comparisons often offer insufficient feedback for effective learning. We introduce RILe (Reinforced Imitation Learning), a framework that combines the strengths of imitation learning and inverse reinforcement learning to learn a dense reward function efficiently and achieve strong performance in high-dimensional tasks. RILe employs a novel trainer-student framework: the trainer learns an adaptive reward function, and the student uses this reward signal to imitate expert behaviors. By dynamically adjusting its guidance as the student evolves, the trainer provides nuanced feedback across different phases of learning. Our framework produces high-performing policies in high-dimensional tasks where direct imitation fails to replicate complex behaviors. We validate RILe in challenging robotic locomotion tasks, demonstrating that it significantly outperforms existing methods and achieves near-expert performance across multiple settings.

Christoph Gebhardt, Antti Oulasvirta, Otmar Hilliges

How do people decide how long to continue in a task, when to switch, and to which other task? Understanding the mechanisms that underpin task interleaving is a long-standing goal in the cognitive sciences. Prior work suggests greedy heuristics and a policy maximizing the marginal rate of return. However, it is unclear how such a strategy would allow for adaptation to everyday environments that offer multiple tasks with complex switch costs and delayed rewards. Here we develop a hierarchical model of supervisory control driven by reinforcement learning (RL). The supervisory level learns to switch using task-specific approximate utility estimates, which are computed on the lower level. A hierarchically optimal value function decomposition can be learned from experience, even in conditions with multiple tasks and arbitrary and uncertain reward and cost structures. The model reproduces known empirical effects of task interleaving. It yields better predictions of individual-level data than a myopic baseline in a six-task problem (N=211). The results support hierarchical RL as a plausible model of task interleaving.

Christoph Gebhardt, Otmar Hilliges

Tools for quadrotor trajectory design have enabled single videographers to create complex aerial video shots that previously required dedicated hardware and several operators. We build on this prior work by studying film-maker's working practices which informed a system design that brings expert workflows closer to end-users. For this purpose, we propose WYFIWYG, a new quadrotor camera tool which (i) allows to design a video solely via specifying its frames, (ii) encourages the exploration of the scene prior to filming and (iii) allows to continuously frame a camera target according to compositional intentions. Furthermore, we propose extensions to an existing algorithm, generating more intuitive angular camera motions and producing spatially and temporally smooth trajectories. Finally, we conduct a user study where we evaluate how end-users work with current videography tools. We conclude by summarizing the findings of work as implications for the design of UIs and algorithms of quadrotor camera tools.