Abstract This paper proposes a novel knowledge-guided learning model for low-light image enhancement. The developed pipeline improves image visibility by addressing noise and contrast issues, detail preservation, and color balancing. This is achieved by integrating denoising and Retinex techniques with vision transformers. Given as little as a single low-light input image, a set of intermediate exposures are generated by means of gamma transform, which serve as inputs to an ensemble of ten transformer models to produce enhanced outputs. An adaptive exposure selection process is then applied based on a composite image quality score. Finally, the selected outputs are fused in a multi-scale manner using weight maps based on contrast, saturation, and well-exposedness features. Extensive experiments on benchmark datasets, LOL-v1, LOL-v2-Real, LOL-v2-Synthetic, and a unified dataset, demonstrate that the proposed method is competitive with state-of-the-art techniques and shows a significant advantage when processing images captured in extremely low-light conditions. In addition, the developed method is successfully applied to the image dehazing problem without any further optimization.

Stroke survivors’ unsupervised therapeutic exercise motions are often accompanied by harmful compensatory motions that prevent proper motor recovery and introduce additional health issues. These compensatory motions are often performed unconsciously and are difficult to prevent. Motion-based games show promise for motivating patients to perform stroke rehabilitation exercises at home by themselves. Currently, exercises with these games are likely to contain undesired compensatory motions. In this article, we provide the design and empirical evaluation of a motion-based game system that addresses the issue of compensation in therapeutic games. We introduce a technique to identify and measure compensation, develop a game that meaningfully uses exercise and compensation as inputs, and use incentives and disincentives to reduce compensation. We show that this technique outperforms existing approaches by significantly reducing compensatory motions during therapeutic exercise. This has important implications for therapeutic games, which can use our findings to improve the quality of motions to be closer to therapist-supervised motions. Our techniques can increase the effectiveness of therapeutic games and reduce the possibility that they may cause harm in long-term use.

Stroke is the leading cause of long-term disability among adults in industrialized nations; approximately 80% of people who survive a stroke experience motor disabilities. Recovery requires hundreds of daily repetitions of therapeutic exercises, often without therapist supervision. When performing therapy alone, people with limited motion often compensate for the lack of motion in one joint by moving another one. This compensation can impede the recovery progress and create new health problems. In this work we contribute (1) a methodology to reliably sense compensatory torso motion in the context of shoulder exercises done by persons with stroke and (2) the design and experimental evaluation of operant-conditioning-based strategies for games that aim to reduce compensatory torso motion. Our results show that these strategies significantly reduce compensatory motions compared to alternatives.

In the United States alone, more than five million people are living with long term motor impairments caused by a stroke. Video game-based therapies show promise in helping people recover lost range of motion and motor control. While researchers have demonstrated the potential utility of game-based rehabilitation through controlled studies, relatively little work has explored longer-term home-based use of therapeutic games. We conducted a six-week home study with a 62 year old woman who was seventeen years post-stroke. She played therapeutic games for approximately one hour a day, five days a week. Over the six weeks, she recovered significant motor abilities, which is unexpected given the time since her stroke. Through observations and interviews, we present lessons learned about the barriers and opportunities that arise from long-term home-based use of therapeutic games.

Stroke is the leading cause of long term disability among adults in industrialized nations. The partial paralysis that stroke patients often experience can make independent living difficult or impossible. Research suggests that many of these patients could recover by performing hundreds of daily repetitions of motions with their affected limbs. Yet, only 31% of patients perform the exercises recommended by their therapists. Home-based stroke rehabilitation games may help motivate stroke patients to perform the necessary exercises to recover. In this paper, we describe a formative study in which we designed and user tested stroke rehabilitation games with both stroke patients and therapists. We describe the lessons we learned about what makes games useful from a therapeutic point of view.

In this paper, we propose a new approach to wireless sensor network assisted navigation while avoiding moving dangers. Our approach relies on an embedded roadmap in the sensor network that always contains safe paths. The roadmap is adaptive, i.e., it adapts its topology to changing dangers. Mobile robots in the environment use the roadmap to reach their destinations. We evaluated the performance of embedded roadmap both in simulations using realistic conditions and with real hardware. Our results show that the proposed navigation algorithm is better suited for sensor networks than traditional navigation field based algorithms. Our observations suggest that there are two drawbacks of traditional navigation field based algorithms, (i) increased power consumption, (ii) message congestion that can prevent important danger avoidance messages to be received by the robots. In contrast, our approach significantly reduces the number of messages on the network (up to 160 times in some scenarios) while increasing the navigation performance.

Autonomous mobile agent navigation is crucial to many mission-critical applications (e.g., search and rescue missions in a disaster area). In this paper, we present how sensor networks may assist probabilistic roadmap methods (PRMs), a class of efficient navigation algorithms particularly suitable for dynamic environments. A key challenge of applying PRM algorithms in dynamic environment is that they require the spatiotemporal sensing of the environment to solve a given navigation problem. To facilitate navigation, we propose a set of query strategies that allow a mobile agent to periodically collect real-time information (e.g., fire conditions) about the environment through a sensor network. Such strategies include local spatiotemporal query (query of spatial neighborhood), global spatiotemporal query (query of all sensors), and border query (query of the border of danger fields). We investigate the impact of different query strategies through simulations under a set of realistic fire conditions. We also evaluate the feasibility of our approach using a real robot and real motes. Our results demonstrate that (1) spatiotemporal queries from a sensor network result in significantly better navigation performance than traditional approaches based on on-board sensors of a robot; (2) the area of local queries represent a tradeoff between communication cost and navigation performance; (3) through in-network processing our border query strategy achieves the best navigation performance at a small fraction of communication cost compared to global spatiotemporal queries.

Abstract In this paper, we present a technique to automatically synthesize dancing motions for arbitrary songs with dance beats. Our technique is based on analysing a musical tune (can be a song or melody) and synthesizing a motion for the virtual character where the character's movement synchronizes to the musical beats. In order to analyse beats of the tune, we developed a fast algorithm. Our motion synthesis algorithm analyses library of stock motions and generates new sequences of movements that were not described in the library. We show that our motion synthesis algorithm is better than previous dance generation techniques. We also present two algorithms to synchronize dance moves and musical beats: a fast greedy algorithm, and a genetic algorithm. Our experimental results show that we can generate new sequences of dance figures in which the dancer reacts to music and dances in synchronization with the music. Copyright © 2005 John Wiley & Sons, Ltd.