RaVÆn: unsupervised change detection of extreme events using ML on-board satellites.

Applications such as disaster management enormously benefit from rapid availability of satellite observations. Traditionally, data analysis is performed on the ground after being transferred-downlinked-to a ground station. Constraints on the downlink capabilities, both in terms of data volume and timing, therefore heavily affect the response delay of any downstream application. In this paper, we introduce RaVÆn, a lightweight, unsupervised approach for change detection in satellite data based on Variational Auto-Encoders (VAEs), with the specific purpose of on-board deployment. RaVÆn pre-processes the sampled data directly on the satellite and flags changed areas to prioritise for downlink, shortening the response time. We verified the efficacy of our system on a dataset-which we release alongside this publication-composed of time series containing a catastrophic event, demonstrating that RaVÆn outperforms pixel-wise baselines. Finally, we tested our approach on resource-limited hardware for assessing computational and memory limitations, simulating deployment on real hardware.

Self-supervised learning for using overhead imagery as maps in outdoor range sensor localization.

Traditional approaches to outdoor vehicle localization assume a reliable, prior map is available, typically built using the same sensor suite as the on-board sensors used during localization. This work makes a different assumption. It assumes that an overhead image of the workspace is available and utilizes that as a map for use for range-based sensor localization by a vehicle. Here, range-based sensors are radars and lidars. Our motivation is simple, off-the-shelf, publicly available overhead imagery such as Google satellite images can be a ubiquitous, cheap, and powerful tool for vehicle localization when a usable prior sensor map is unavailable, inconvenient, or expensive. The challenge to be addressed is that overhead images are clearly not directly comparable to data from ground range sensors because of their starkly different modalities. We present a learned metric localization method that not only handles the modality difference, but is also cheap to train, learning in a self-supervised fashion without requiring metrically accurate ground truth. By evaluating across multiple real-world datasets, we demonstrate the robustness and versatility of our method for various sensor configurations in cross-modality localization, achieving localization errors on-par with a prior supervised approach while requiring no pixel-wise aligned ground truth for supervision at training. We pay particular attention to the use of millimeter-wave radar, which, owing to its complex interaction with the scene and its immunity to weather and lighting conditions, makes for a compelling and valuable use case.

kRadar++: Coarse-to-Fine FMCW Scanning Radar Localisation.

This paper presents a novel two-stage system which integrates topological localisation candidates from a radar-only place recognition system with precise pose estimation using spectral landmark-based techniques. We prove that the-recently available-seminal radar place recognition (RPR) and scan matching sub-systems are complementary in a style reminiscent of the mapping and localisation systems underpinning visual teach-and-repeat (VTR) systems which have been exhibited robustly in the last decade. Offline experiments are conducted on the most extensive radar-focused urban autonomy dataset available to the community with performance comparing favourably with and even rivalling alternative state-of-the-art radar localisation systems. Specifically, we show the long-term durability of the approach and of the sensing technology itself to autonomous navigation. We suggest a range of sensible methods of tuning the system, all of which are suitable for online operation. For both tuning regimes, we achieve, over the course of a month of localisation trials against a single static map, high recalls at high precision, and much reduced variance in erroneous metric pose estimation. As such, this work is a necessary first step towards a radar teach-and-repeat (RTR) system and the enablement of autonomy across extreme changes in appearance or inclement conditions.

Empowering phase II clinical trials to reduce phase III failures.

The large number of failures in phase III clinical trials, which occur at a rate of approximately 45%, is studied herein relative to possible countermeasures. First, the phenomenon of failures is numerically described. Second, the main reasons for failures are reported, together with some generic improvements suggested in the related literature. This study shows how statistics explain, but do not justify, the high failure rate observed. The rate of failures due to a lack of efficacy that are not expected, is considered to be at least 10%. Expanding phase II is the simplest and most intuitive way to reduce phase III failures since it can reduce phase III false negative findings and launches of phase III trials when the treatment is positive but suboptimal. Moreover, phase II enlargement is discussed using an economic profile. As resources for research are often limited, enlarging phase II should be evaluated on a case-by-case basis. Alternative strategies, such as biomarker-based enrichments and adaptive designs, may aid in reducing failures. However, these strategies also have very low application rates with little likelihood of rapid growth.

Computing individual and collective ethical utility for optimally planning phase III trials.

A quantitative evaluation of individual and collective ethics is proposed here, with the aim of providing a tool for sample size determination/estimation that goes further than the standard power setting of 80-90%. Individual ethics deal with issues that concern the patients enrolled in the trial, where collective ones concern the patients not enrolled in the trial, and who might benefit from a positive result. The global ethical utility (GEU) of a phase III trial is introduced here, being the summation of individual and collective ethical utilities, and can be viewed as a function of the sample size. The GEU model is based on the extent of the efficacy of the treatments in study, of the quality of life of the patients being treated, of the effects of potential adverse reactions, it accounts for the duration of the periods of interest and for the size of population groups, and also embeds the experimental power. This work aims at arguing the case for GEU adoption for sample size determination. The sample size that maximizes GEU can be adopted for planning the trial, even when providing a power value out of the classical range [.8,.9]. Alternatively, among the sample sizes based on power values of 80% and 90%, the one providing the highest GEU can be adopted. Intuitively, when a treatment is assumed to work well, to have few adverse effects, and is expected to improve the QoL of the ill population for a considerable amount of time, collective ethics may prevail giving ethically optimal sample sizes larger than usual, and consequent quite high power values (e.g. 99%). Instead, medium, though still clinically meaningful, levels of effect, considerable adverse reactions, and limited life expectation and QoL improvement, might shift the ethical balance on individual ethics and give an ethically optimal sample size providing a power lower than standard values (e.g. 70%). Some examples and an application in the cardiovascular area, including sensitivity analyses of the results based on the so-called Bayesian "assurance" technique, are also discussed. Several possible extensions of the model related to particular clinical frameworks are also presented.

Profit Evaluations When Adaptation by Design Is Applied.

BACKGROUND: Adaptation by design consists in conservatively estimating the phase III sample size on the basis of phase II data; it is also called conservative sample size estimation (CSSE). The usual assumptions are that the effect size is the same in both phases and that phase II data are not used for phase III confirmatory analysis. CSSE has been introduced to increase the rate of successful trials, and it can be applied in most clinical areas. CSSE reduces the probability of underpowered experiments and can improve the overall success probability of phase II and III, but it also increases phase III sample size, increasing the time and cost of experiments. Thus, the balance between higher revenue and greater cost is the issue. METHODS: A profit model was built assuming that CSSE was applied and considering income per patient, annual incidence, time on market, market share, phase III success probability, fixed cost of the 2 phases, and cost per patient under treatment. RESULTS: Profit turns out to be a random variable depending on phase II sample size and conservativeness. Profit moments are obtained in a closed formula. Profit utility, which is a linear function of profit expectation and volatility, is evaluated in accordance with the modern theory of investment performances. Indications regarding phase II sample size and conservativeness can be derived on the basis of utility, for example, through utility optimization. CONCLUSIONS: CSSE can be adopted in many different statistical problems, and consequently the profit evaluations proposed here can be widely applied.

Stability criteria for the outcomes of statistical tests to assess drug effectiveness with a single study.

At least two adequate and well-controlled clinical studies are usually required to support effectiveness of a certain treatment. In some circumstances, however, a single study providing strong results may be sufficient. Some statistical stability criteria for assessing whether a single study provides very persuasive results are known. A new criterion is introduced, and it is based on the conservative estimation of the reproducibility probability in addition to the possibility of performing statistical tests by referring directly to the reproducibility probability estimate. These stability criteria are compared numerically and conceptually. This work aims to help both regulatory agencies and pharmaceutical companies to decide if the results of a single study may be sufficient to establish effectiveness.

Conservative sample size estimation in nonparametrics.

Due to the uncertainty of the results of phase II trials, underpowered phase III trials are often planned. In recent literature the conservative approach for sample size estimation was proposed. Some authors, in the parametric framework, make use of the lower bound of the effect size for conservatively estimating the true power, and so the sample sizes. Here, we present a general bootstrap method for conservatively estimating, on the basis of phase II data, the sample size needed for a phase III trial. The method we propose is based on the use of nonparametric lower bounds for the true power of the test. A wide study is shown for comparing the performances of the new method in estimating the power of the Wilcoxon rank-sum test with those given by standard techniques based on the asymptotic normality of the test statistic. Results indicate that when the phase II sample size is around the ideal sample size for the phase III, the bootstrap provides better results than the other techniques. Since the method is general, it could be used for planning clinical trials for testing superiority, for testing noninferiority, and for more complicated situations, e.g., for testing multiple endpoints.