Ensemble Foreground Management for Unsupervised Object Discovery

Unsupervised object discovery (UOD) aims to detect and segment objects in 2D images without handcrafted annotations. Recent progress in self-supervised representation learning has led to some success in UOD algorithms. However, the absence of ground truth provides existing UOD methods with two challenges: 1) determining if a discovered region is foreground or background, and 2) knowing how many objects remain undiscovered. To address these two problems, previous solutions rely on foreground priors to distinguish if the discovered region is foreground, and conduct one or fixed iterations of discovery. However, the existing foreground priors are heuristic and not always robust, and a fixed number of discoveries leads to under or over-segmentation, since the number of objects in images varies. This paper introduces UnionCut, a robust and well-grounded foreground prior based on min-cut and ensemble methods that detects the union of foreground areas of an image, allowing UOD algorithms to identify foreground objects and stop discovery once the majority of the foreground union in the image is segmented. In addition, we propose UnionSeg, a distilled transformer of UnionCut that outputs the foreground union more efficiently and accurately. Our experiments show that by combining with UnionCut or UnionSeg, previous state-of-the-art UOD methods witness an increase in the performance of single object discovery, saliency detection and self-supervised instance segmentation on various benchmarks. The code is available at https://github.com/YFaris/UnionCut.

Tangles: Unpacking Extended Collision Experiences with Soma Trajectories

We reappraise the idea of colliding with robots, moving from a position that tries to avoid or mitigate collisions to one that considers them an important facet of human interaction. We report on a soma design workshop that explored how our bodies could collide with telepresence robots, mobility aids, and a quadruped robot. Based on our findings, we employed soma trajectories to analyse collisions as extended experiences that negotiate key transitions of consent, preparation, launch, contact, ripple, sting, untangle, debris and reflect. We then employed these ideas to analyse two collision experiences, an accidental collision between a person and a drone, and the deliberate design of a robot to play with cats, revealing how real-world collisions involve the complex and ongoing entanglement of soma trajectories. We discuss how viewing collisions as entangled trajectories, or tangles, can be used analytically, as a design approach, and as a lens to broach ethical complexity.

Vibrotactile information coding strategies for a body-worn vest to aid robot-human collaboration

This paper explores the use of a body-worn vibrotactile vest to convey real-time information from robot to operator. Vibrotactile communication could be useful in providing information without compropmising or loading a person's visual or auditory perception. This paper considers applications in Urban Search and Rescue (USAR) scenarios where a human working alongside a robot is likely to be operating in high cognitive load conditions. The focus is on understanding how best to convey information considering different vibrotactile information coding strategies to enhance scene understanding in scenarios where a robot might be operating remotely as a scout. In exploring information representation, this paper introduces Semantic Haptics, using shapes and patterns to represent certain events as if the skin was a screen, and shows how these lead to bettter learnability and interpreation accuracy.

The Dilemma of Decision-Making in the Real World: When Robots Struggle to Make Choices Due to Situational Constraints

In order to demonstrate the limitations of assistive robotic capabilities in noisy real-world environments, we propose a Decision-Making Scenario analysis approach that examines the challenges due to user and environmental uncertainty, and incorporates these into user studies. The scenarios highlight how personalization can be achieved through more human-robot collaboration, particularly in relation to individuals with visual, physical, cognitive, auditory impairments, clinical needs, environmental factors (noise, light levels, clutter), and daily living activities. Our goal is for this contribution to prompt reflection and aid in the design of improved robots (embodiment, sensors, actuation, cognition) and their behavior, and we aim to introduces a groundbreaking strategy to enhance human-robot collaboration, addressing the complexities of decision-making under uncertainty through a Scenario analysis approach. By emphasizing user-centered design principles and offering actionable solutions to real-world challenges, this work aims to identify key decision-making challenges and propose potential solutions.