Joint Dynamic Pose Image and Space Time Reversal for Human Action Recognition from Videos

Human action recognition aims to classify a given video according to which type of action it contains. Disturbance brought by clutter background and unrelated motions makes the task challenging for video frame-based methods. To solve this problem, this paper takes advantage of pose estimation to enhance the performances of video frame features. First, we present a pose feature called dynamic pose image (DPI), which describes human action as the aggregation of a sequence of joint estimation maps. Different from traditional pose features using sole joints, DPI suffers less from disturbance and provides richer information about human body shape and movements. Second, we present attention-based dynamic texture images (att-DTIs) as pose-guided video frame feature. Specifically, a video is treated as a space-time volume, and DTIs are obtained by observing the volume from different views. To alleviate the effect of disturbance on DTIs, we accumulate joint estimation maps as attention map, and extend DTIs to attention-based DTIs (att-DTIs). Finally, we fuse DPI and att-DTIs with multi-stream deep neural networks and late fusion scheme for action recognition. Experiments on NTU RGB+D, UTD-MHAD, and Penn-Action datasets show the effectiveness of DPI and att-DTIs, as well as the complementary property between them.

Power-complementary IIR filter pairs with an adjustable crossover frequency

This paper introduces two classes of power-complementary recursive low-pass/high-pass filter pairs with an adjustable crossover frequency in such a way that the stopband attenuation of both filters remains the same. For each class, the filter pair is constructed using two all-pass sub-filters as building blocks. Based on the properties elliptic minimal Q-factors transfer functions, simple expressions are derived for evaluating the coefficients in all-pass sections in order to achieve the desired crossover frequency. The design procedures are developed for synthesizing power-complementary filter pairs implemented as a parallel connection of two all-pass sub-filters and for the tapped cascaded interconnections of two identical all-pass sub-filters. The direct parallel connection has both the power-complementary and all-pass complementary property. The second class of filters constructed using several identical copies of the two all-pass filters possesses the power-complementary property.

THE INTERFERENCE OF INACTIVE SERUM AND EGG-WHITE IN THE PHENOMENON OF COMPLEMENT FIXATIONS

The fixing property of a specific precipitate and of syphilitic serum in the presence of certain antigenic lipoids, can be removed by adding certain non-complementary proteins of blood serum or hen's egg. This disappearance of the complementary activity in the syphilis reaction, as well as in the true Bordet-Gengou reaction, is a phenomenon which incidentally accompanies the fixation of certain serum constituents, some of which possess a complementary activity. The presence or absence of the complementary property in these protein components does not influence fixation. Whether the disappearance of the complementary activity during the phenomenon of so-called fixation is due to a mechanical precipitation of the molecules through absorption or whether it is due to a physico-chemical alteration of the active molecules, is unknown. It is more probable that a chemical interaction takes place in the case of the syphilis reaction. Certain sera, for example, those derived from man and goat, show a low fixability. It is interesting to note that the fixability is gradually diminished when these sera and egg-white are heated to a temperature above 56° C., and totally disappears at 85° C. The coagulation of proteins with absolute alcohol or by boiling, destroys their interfering property. The fact that the fixation is not selectively directed towards complement, has a very important meaning for exact serology. The one-sided accuracy as to the complementary unity is no longer sufficient for quantitative work. Both the complementary and the volumetric unity of a serum serving as the source of complement should be taken into consideration. Besides, the fixability of the sera of various species of animals must also be considered. From these facts a formula may be derived for deciding the degree of suitableness of a serum. see PDF for Equation X is the degree of suitableness; K, the species constant for the fixability; P, the complementary activity; and V, the volume of serum. It will be seen that the suitableness is proportional to the fixability constant and the complementary unity, and inversely proportional to the volume of serum employed. As to what species yields the largest value for X, we refer the reader to our studies published elsewhere.