Recently, convolutional neural networks with 3D kernels (3D CNNs) have been very popular in computer vision community as a result of their superior ability of extracting spatio-temporal features within video frames compared to 2D CNNs. Although there has been great advances recently to build resource efficient 2D CNN architectures considering memory and power budget, there is hardly any similar resource efficient architectures for 3D CNNs. In this paper, we have converted various well-known resource efficient 2D CNNs to 3D CNNs and evaluated their performance on three major benchmarks in terms of classification accuracy for different complexity levels. We have experimented on (1) Kinetics-600 dataset to inspect their capacity to learn, (2) Jester dataset to inspect their ability to capture motion patterns, and (3) UCF-101 to inspect the applicability of transfer learning. We have evaluated the run-time performance of each model on a single Titan XP GPU and a Jetson TX2 embedded system. The results of this study show that these models can be utilized for different types of real-world applications since they provide real-time performance with considerable accuracies and memory usage. Our analysis on different complexity levels shows that the resource efficient 3D CNNs should not be designed too shallow or narrow in order to save complexity. The codes and pretrained models used in this work are publicly available 1 .

This study aimed to determine the effect of accelerated weathering on gloss, surface hardness and colour changes of Scots pine (Pinus sylvestris L.). Test samples were impregnated with Adolit KD‐5, Wolmanit CX‐8 and Celcure AC‐500 covered with cellulosic and polyurethane varnishes. The results showed that the values of surface hardness and gloss increased after accelerated weathering. While the surface hardness of Scots pine was increased for impregnated and polyurethane‐coated varnish, it decreased for impregnated and cellulosic varnish‐coated Scots pine after 1000 hours of accelerated weathering exposure. Copper‐based chemical impregnation and varnish coating developed the gloss of Scots pine specimens relative to the surface characteristics observed in single‐coated Scots pine specimens. While the most appropriate chemical was Celcure AC‐500 for surface hardness, it was Adolit KD‐5 for the gloss of Scots pine after 1000 hours of accelerated weathering exposure. Wood specimens impregnated prior to the application of varnish were more effective in stabilising the colour of Scots pine than Scots pine only coated with varnish. Polyurethane varnish‐treated Scots pine showed better colour stability for each partial and total accelerated weathering exposure period. The total colour changes were lowest for polyurethane varnish‐coated Scots pine impregnated with Celcure AC‐500 after 1000 hours of accelerated weathering exposure.

Real-time recognition of dynamic hand gestures from video streams is a challenging task since (i) there is no indication when a gesture starts and ends in the video, (ii) performed gestures should only be recognized once, and (iii) the entire architecture should be designed considering the memory and power budget. In this work, we address these challenges by proposing a hierarchical structure enabling offline-working convolutional neural network (CNN) architectures to operate online efficiently by using sliding window approach. The proposed architecture consists of two models: (1) A detector which is a lightweight CNN architecture to detect gestures and (2) a classifier which is a deep CNN to classify the detected gestures. In order to evaluate the single-time activations of the detected gestures, we propose to use Levenshtein distance as an evaluation metric since it can measure misclassifications, multiple detections, and missing detections at the same time. We evaluate our architecture on two publicly available datasets -EgoGesture and NVIDIA Dynamic Hand Gesture Datasets -which require temporal detection and classification of the performed hand gestures. ResNeXt-101 model, which is used as a classifier, achieves the state-of-the-art offline classification accuracy of 94.04% and 83.82% for depth modality on EgoGesture and NVIDIA benchmarks, respectively. In real-time detection and classification, we obtain considerable early detections while achieving performances close to offline operation. The codes and pretrained models used in this work are publicly available 1 .

Weathering performance of impregnated and coated wood products is an important issue that influences their appearance as well as their service life after outdoor or indoor exposure. A novel procedure to improve the weathering performance of Scots pine wood (Pinus sylvestris L.) is proposed in this study. Wood samples were impregnated with ammonium tetrafluoroborate (ATFB), ammonium pentaborate (APB) and boric acid (BA), and afterward coated with a layer of liquid glass (LG). Surface hardness, color changes, and surface roughness of wood samples were investigated after 250 h and 500 h of exposure to artificial weathering. The results revealed that, except for untreated (control), all other treatment groups caused an increase in surface hardness of Scots pine after weathering. A decrease in the CIE L* value corresponds to the darkening of samples after weathering. Borates-impregnated and LG-coated Scots pine samples possessed the highest color stability. However, all treatment combinations resulted in reddish and yellowish tones after weathering. Exposure to weathering increased surface roughness of wood samples. The surface roughness of the untreated (control) group was higher than that of impregnated and coated samples.

In this study, it was aimed to investigate the effects of weathering on some surface characteristics such as color and surface roughness changes of Scots pine impregnated with copper-containing chemical such as Wolmanit CX-8 (WCX-8) and varnished with synthetic varnish (SV), cellulosic varnish (CV), and polyurethane varnish (PV) were investigated. Results showed that while the WCX-8 impregnated and PV coated Scots pine specimens showed better color stability than other treatment groups after weathering, only CV coated Scots pine gave the most negative effect on color stability. While, the untreated (control) wood surface turned from red to green and yellow to blue respectively, after weathering, other all treatment groups gave reddish and yellowish tone after weathering. Weathering conditions increased the surface roughness of control (untreated) and other all treatment groups. The control group gave a rougher surface than other treatment groups after weathering. Surface roughness increases were the lower for CV coated Scots pine wood than other treatment groups. The results showed that while WCX-8 impregnation before varnishing gave better color characteristics, generally it caused to increase the surface roughness of Scots pine after weathering.