She adapts to her student: An expert pragmatic speaker tailoring her referring expressions to the Layman listener

Communication is a dynamic process through which interlocutors adapt to each other. In the development of conversational agents, this core aspect has been put aside for several years since the main challenge was to obtain conversational neural models able to produce utterances and dialogues that at least at the surface level are human-like. Now that this milestone has been achieved, the importance of paying attention to the dynamic and adaptive interactive aspects of language has been advocated in several position papers. In this paper, we focus on how a Speaker adapts to an interlocutor with different background knowledge. Our models undergo a pre-training phase, through which they acquire grounded knowledge by learning to describe an image, and an adaptive phase through which a Speaker and a Listener play a repeated reference game. Using a similar setting, previous studies focus on how conversational models create new conventions; we are interested, instead, in studying whether the Speaker learns from the Listener's mistakes to adapt to his background knowledge. We evaluate models based on Rational Speech Act (RSA), a likelihood loss, and a combination of the two. We show that RSA could indeed work as a backbone to drive the Speaker toward the Listener: in the combined model, apart from the improved Listener's accuracy, the language generated by the Speaker features the changes that signal adaptation to the Listener's background knowledge. Specifically, captions to unknown object categories contain more adjectives and less direct reference to the unknown objects.

An adaptive model of health system organization and responses helped Vietnam to successfully halt the Covid-19 pandemic: What lessons can be learned from a resource-constrained country

Coping with the COVID‐19 pandemic has been painful and no single model for such a purpose is perfect. However, sharing experiences is the best way for countries to learn real‐time lessons and adapt to this rapidly changing pandemic. This commentary shares with the international community how an adaptive model of health system organization and responses helped Vietnam to break transmission of coronavirus. We find that an effective model is adaptive to time and context, and mobilizes and engages the wider society. We identify merging of different health system units into Center for Diseases Controls as a health system organization that saved massive resources. The early establishment of a formal committee responding to the pandemic helped unify every public health strategy. The mobilization of different stakeholders and communities added resources and facilitated a synchronous implementation of response strategies, even where those strategies involved significant personal or financial sacrifice. National training on Covid‐19 treatment for healthcare professionals across the entire hospital system was useful to expand the health service availability. Quickly published response guidelines helped to activate every level of the health system and involve every sector of society. A strategy of keeping high alert and preemptive action is also essential for coping with the pandemic.

A Model to Support Fluid Transitions between Environments for Mobile Augmented Reality Applications

The adaptability between different environments remains a challenge for Mobile Augmented Reality (MAR). If not done seamlessly, such transitions may cause discontinuities in navigation, consequently disorienting users and undermining the acceptance of this technology. The transition between environments is hard because there are currently no localization techniques that work well in any place: sensor-based applications can be harmed by obstacles that hamper sensor communication (e.g., GPS) and by infrastructure limitations (e.g., Wi-Fi), and image-based applications can be affected by lighting conditions that impair computer vision techniques. Hence, this paper presents an adaptive model to perform transitions between different types of environments for MAR applications. The model has a hybrid approach, choosing the best combination of long-range sensors, short-range sensors, and computer vision techniques to perform fluid transitions between environments that mitigate problems in location, orientation, and registration. To assess the model, we developed a MAR application and conducted a navigation test with volunteers to validate transitions between outdoor and indoor environments, followed by a short interview. The results show that the transitions were well succeeded, since the application self-adapted to the studied environments, seamlessly changing sensors when needed.

Improved Point-Line Feature Based Visual SLAM Method for Indoor Scenes

In the study of indoor simultaneous localization and mapping (SLAM) problems using a stereo camera, two types of primary features-point and line segments-have been widely used to calculate the pose of the camera. However, many feature-based SLAM systems are not robust when the camera moves sharply or turns too quickly. In this paper, an improved indoor visual SLAM method to better utilize the advantages of point and line segment features and achieve robust results in difficult environments is proposed. First, point and line segment features are automatically extracted and matched to build two kinds of projection models. Subsequently, for the optimization problem of line segment features, we add minimization of angle observation in addition to the traditional re-projection error of endpoints. Finally, our model of motion estimation, which is adaptive to the motion state of the camera, is applied to build a new combinational Hessian matrix and gradient vector for iterated pose estimation. Furthermore, our proposal has been tested on EuRoC MAV datasets and sequence images captured with our stereo camera. The experimental results demonstrate the effectiveness of our improved point-line feature based visual SLAM method in improving localization accuracy when the camera moves with rapid rotation or violent fluctuation.

Influence of combined iron supplementation and simulated hypoxia on the haematological module of the athlete biological passport

The integrity of the athlete biological passport (ABP) is underpinned by understanding normal fluctuations of its biomarkers to environmental or medical conditions, for example, altitude training or iron deficiency. The combined impact of altitude and iron supplementation on the ABP was evaluated in endurance-trained athletes (n = 34) undertaking 3 weeks of simulated live-high: train-low (14 h.d^-1 , 3000 m). Athletes received either oral, intravenous (IV) or placebo iron supplementation, commencing 2 weeks prior and continuing throughout hypoxic exposure. Venous blood was sampled twice prior, weekly during, and up to 6 weeks after altitude. Individual ABP thresholds for haemoglobin concentration ([Hb]), reticulocyte percentage (%retic), and OFF score were calculated using the adaptive model and assessed at 99% and 99.9% specificity. Eleven athletes returned values outside of the calculated reference ranges at 99%, with 8 at 99.9%. The percentage of athletes exceeding the thresholds in each group was similar, but IV returned the most individual occurrences. A similar frequency of abnormalities occurred across the 3 biomarkers, with abnormal [Hb] and OFF score values arising mainly during-, and %retic values mainly post- altitude. Removing samples collected during altitude from the model resulted in 10 athletes returning abnormal values at 99% specificity, 2 of whom had not triggered the model previously. In summary, the abnormalities observed in response to iron supplementation and hypoxia were not systematic and mostly in line with expected physiological adaptations. They do not represent a uniform weakness in the ABP. Nevertheless, altitude training and iron supplementation should be carefully considered by experts evaluating abnormal ABP profiles.