The zebrafish offers an excellent compromise between system complexity and practical simplicity and has been suggested as a translational research tool for the analysis of human brain disorders associated with abnormalities of social behavior. Unlike laboratory rodents zebrafish are diurnal, thus visual cues may be easily utilized in the analysis of their behavior and brain function. Visual cues, including the sight of conspecifics, have been employed to induce social behavior in zebrafish. However, the method of presentation of these cues and the question of whether computer animated images versus live stimulus fish have differential effects have not been systematically analyzed. Here, we compare the effects of five stimulus presentation types: live conspecifics in the experimental tank or outside the tank, playback of video-recorded live conspecifics, computer animated images of conspecifics presented by two software applications, the previously employed General Fish Animator, and a new application Zebrafish Presenter. We report that all stimuli were equally effective and induced a robust social response (shoaling) manifesting as reduced distance between stimulus and experimental fish. We conclude that presentation of live stimulus fish, or 3D images, is not required and 2D computer animated images are sufficient to induce robust and consistent social behavioral responses in zebrafish.
One important aspect of effective human--robot collaborations is the ability for robots to adapt quickly to the needs of humans. While techniques like deep reinforcement learning have demonstrated success as sophisticated tools for learning robot policies, the fluency of human-robot collaborations is often limited by these policies' inability to integrate changes to a user's preferences for the task. To address these shortcomings, we propose a novel approach that can modify learned policies at execution time via symbolic if-this-then-that rules corresponding to a modular and superimposable set of low-level constraints on the robot's policy. These rules, which we call Transparent Matrix Overlays, function not only as succinct and explainable descriptions of the robot's current strategy but also as an interface by which a human collaborator can easily alter a robot's policy via verbal commands. We demonstrate the efficacy of this approach on a series of proof-of-concept cooking tasks performed in simulation and on a physical robot.