RESUMEN
In real-world applications, robotic systems collect vast amounts of new data from ever-changing environments over time. They need to continually interact and learn new knowledge from the external world to adapt to the environment. Particularly, lifelong object recognition in an online and interactive manner is a crucial and fundamental capability for robotic systems. To meet this realistic demand, in this article, we propose an online active continual learning (OACL) framework for robotic lifelong object recognition, in the scenario of both classes and domains changing with dynamic environments. First, to reduce the labeling cost as much as possible while maximizing the performance, a new online active learning (OAL) strategy is designed by taking both the uncertainty and diversity of samples into account to protect the information volume and distribution of data. In addition, to prevent catastrophic forgetting and reduce memory costs, a novel online continual learning (OCL) algorithm is proposed based on the deep feature semantic augmentation and a new loss-based deep model and replay buffer update, which can mitigate the class imbalance between the old and new classes and alleviate confusion between two similar classes. Moreover, the mistake bound of the proposed method is analyzed in theory. OACL allows robots to select the most representative new samples to query labels and continually learn new objects and new variants of previously learned objects from a nonindependent and identically distributed (i.i.d.) data stream without catastrophic forgetting. Extensive experiments conducted on real lifelong robotic vision datasets demonstrate that our algorithm, even trained with fewer labeled samples and replay exemplars, can achieve state-of-the-art performance on OCL tasks.
RESUMEN
Extractable trace level lead in artificial sweat solution from ecological textiles is a key item limited by eco-textile standard. But the content of this extractable Pb is not so easy to determine for the strict limit of eco-textile standard, the complicatedness of extractable solution matrix and the strong background interference of NaCl. In the present paper a method for the determination of trace extractable lead in artificial acid sweat from ecological textiles by graphite furnace atomic absorption spectrometry (GFAAS) is described. Based on a number of experiments by using different single and mixed matrix modifiers including (NH4)2 H2PO4, NH4 NO3, Pd(NO3)2, Ni(NO3)2 and ascorbic acid, an effective modifier and its quantity were selected and the graphite furnace operating parameters were optimized. Experimental test results revealed that adding 5 mL (1 : 1) mixed solution of 50 g x L(-1) ammonium nitrate and 100 mg x L(-1) palladium regent was an effective way to inhibit volatile lead and reduce background signals. The detection limit could reach a low level of 0.7 microg x L(-1). The relative standard deviation was 3.2%. Under the optimum experimental conditions, the recoveries ranged between 95.5% and 105%.