Search

Recognizing human-object interaction (HOI) is an important research topic in computer vision. With the great success of deep learning in image classification, the HOI recognition task has also made great progress. However, the problems of instance imbalance and combinatorial explosion still remain the key challenges, which restrict the performance of HOI recognition methods. Therefore, this paper formulates HOI recognition in a few-shot scene to tackle the above problems and proposes a novel task-aware dual prototypical network (TDP-Net) to address few-shot HOI task. Specifically, it first assigns semantic-aware task representations for different tasks as their prior knowledge, subsequently generates attention weights by semantic graph attention module (SGA-Module). It effectively weights the importance on different regions of the visual features, adaptively for different task conditions, which realizes to reason for novel tasks. In addition, it designs a dual prototypes module (DP-Module) to generate both action class prototypes and object class prototypes, which classifies the verb and noun labels respectively. The complex visual relationships between actions and objects can be effectively separated by constructing class prototypes for actions and objects. Meanwhile, owing to the similarity among the related interactions, the knowledge is transferred to the new interactions by reorganizing the action and object prototypes. The experimental results show that the average accuracies of this model outperform the baseline by 3.2 percentage points and 15.7 percentage points on two exper-imental settings, which verifies its effectiveness on the few-shot HOI task.

The technology for flue gas desulfurization (FGD) with seawater is widely adopted by coal-fired power plants in coastal areas. SO2 in the flue gas is absorbed by alkaline seawater and transfered in aqueous phase as sulfite (SO3(2-)), and most SO3(2-) is transformed to sulfate (SO4(2-)) after an aeration process. The remaining SO3(2-) in the seawater discharged to sea area may be harmful to marine organism because of its biological toxicity, thus it is necessary to determine the concentration of SO3(2-) in the seawater for desulfurization. In this study, the method of determination of SO3(2-) in the seawater by ion chromatography was investigated. The separation was achieved on an IonPac AS14A column with 14 mmol/L NaOH-12 mmol/L Na2 CO3 solution as the mobile phase at a flow rate of 1.2 mL/min, and the detection was performed by a pulsed amperometric detector. Formaldehyde was added as a protective agent when sampling because the SO3(2-) is easy to be oxidized. To avoid the formation of Mg (OH)2 in the mobile phase with high pH value which might block the column, the Mg(2+) in seawater was precipitated by NaOH solution (pH 12.0) before sample determination. The method showed good linearity within the range of 0-100 mg/L with an average recovery of 116.8%. The method detection limit (MDL) reached as low as 0.05 mg/L. The relative standard deviations (RSD) for seawater matrix samples spiked at levels of 7.5, 25.0 and 75.0 mg/L were 2.1% 3.1% and 4.0% (n = 9) , respectively. The method has been applied for the determination of SO3(2-) in flue gas desulfurization seawater with the advantages of being fast, sensitive and selective.