Alpha Fusion Adversarial Attack Analysis Using Deep Learning

The deep learning model encompasses a powerful learning ability that integrates the feature extraction, and classification method to improve accuracy. Convolutional Neural Networks (CNN) perform well in machine learning and image processing tasks like segmentation, classification, detection, identification, etc. The CNN models are still sensitive to noise and attack. The smallest change in training images as in an adversarial attack can greatly decrease the accuracy of the CNN model. This paper presents an alpha fusion attack analysis and generates defense against adversarial attacks. The proposed work is divided into three phases: firstly, an MLSTM-based CNN classification model is developed for classifying COVID-CT images. Secondly, an alpha fusion attack is generated to fool the classification model. The alpha fusion attack is tested in the last phase on a modified LSTM-based CNN (CNN-MLSTM) model and other pre-trained models. The results of CNN models show that the accuracy of these models dropped greatly after the alpha-fusion attack. The highest F1 score before the attack was achieved is 97.45 And after the attack lowest F1 score recorded is 22%. Results elucidate the performance in terms of accuracy, precision, F1 score and Recall.     

Spatial orientation in humans: perception of angular whole-body displacements in two-dimensional trajectories

Vestibular perception of whole-body passive rotation in the horizontal plane was studied by applying two-dimensional (2D) motion to eight blindfolded healthy volunteers: pure rotations in place, corner-like trajectories and arcs of a circular trajectory were randomly applied by means of a remotely controlled robot. Angles embedded in the 2D trajectories were 45 degrees, 90 degrees, 135 degrees and 180 degrees. Stimulation of semicircular canals was the same for all trajectories but was accompanied by concurrent otolith stimulation during circular motion. Subjects participated in two successive experimental sessions. In the first session they were instructed to use a pointer to reproduce the total angular displacement after the motion (REPRODUCTION); in the second session they had to keep pointing towards a remote (15 m) memorised target during the motion (TRACKING). In REPRODUCTION subjects tended to overestimate their rotation angle by 28 +/- 11% (mean +/- SD). There was no systematic effect of the trajectory. Overestimation also occurred when subjects were required to rotate in darkness by 180 degrees (by controlling a joystick). In TRACKING there was virtually no overestimation (6 +/- 17%) and the movement of the pointer matched the dynamics of angular motion. We conclude that (a) the brain can separate and memorise the angular component of complex 2D motion; however, a large inter-individual variability in estimating its amplitude exists; (b) in the range of linear accelerations used in the study, no appreciable effect of otolith-canal perceptual interaction was shown; (c) angular displacements can be dynamically transformed into matched pointing movements; (d) overestimation seems to be typical of delayed judgements of angular displacement and of self-controlled rotations in place. This could be due to the characteristics of the physiological calibration of the vestibular input.     

Enhanced acetylcholine release from cells that have more 15-kDa proteolipid in their membrane, a constituent V-ATPase, and mediatophore

Mediatophore is a protein that translocates acetylcholine (ACh) on calcium action. It is a homopolymer of a 15-kDa proteolipid that is also a constituent of the membrane sector of vacuolar H+-adenosine trisphosphatase (V-ATPase; vacuolar proton pump). Experiments on neuroblastoma cell lines (N18TG-2) that are deficient for ACh release and on cells that are competent for release, such as the glioma C6BU-1 or the N18TG-2/C6BU-1 fusion product NG108-15, show that there is a correlation between ACh release and the 15-kDa proteolipid content of the cell membrane. In another cell line, L-M(TK-), it has been possible to up-regulate ACh release and the membrane proteolipid content after treating the cells with dibutyryl-cyclic AMP or dexamethasone. As mediatophore translocates ACh and as V-ATPase may help vesicular ACh storage, it was interesting to determine the respective role of the two proteins in the observed correlation between release and proteolipid content. After blocking vesicular loading with vesamicol, we did not affect release from these cells, suggesting that the observed correlation may be attributed to mediatophore. The acquisition of an ACh release mechanism would then depend on the process that guides the proteolipid to the plasma membrane of the cell.     

The house as an expression of social worlds: Irbid's elite and their architecture

This paper shows howarchitecture can be explained through the ideaof the Social World. The social world isan abstract concept defined as a cultural area,the boundaries of which are defined by thelimit of effective communication. Within thesocial world there are special norms and valuesand a common outlook towards life. Itencompasses factors like identity, culture,social communication and economic and politicalconditions without being limited to one singlefactor. The paper establishes a link betweenthe contemporary villas of Irbid, the secondlargest city in Jordan, and the social worldsof their inhabitants. The paper focuses onIrbid's University District (UD)differentiating between two groups of socialworlds, the old elite and the new elite. Itshows how the meaning of the house is differentamong the two groups. While the villa among theold elite is treated as a functional object forliving, it is now dealt with as a statusobject – an object for displaying theself – among the new elite.     

Calcium-dependent release specificities of various cell lines loaded with different transmitters

By loading cells in culture with acetylcholine (ACh) we have characterized a calcium-dependent release mechanism and shown that it was expressed independently of synthesis or storage of ACh. (Isra?l et al., 1994, Neurochemistry International 37, 1475-1483; Falk-Vairant et al., 1996a, Proc. Natl. Acad. Sci. U.S.A. 93, 5203-5207; Falk-Vairant et al., 1996b, Neuroscience 75, 353-360; Falk-Vairant et al., 1996c, Journal of Neuroscience Research 45, 195-201). The transmitter loading procedure was applied to two other transmitters, gamma-aminobutyric acid (GABA) and glutamate (Glu). We could then study the specificity of the release mechanism for the three transmitters in a variety of cell lines, including neural-derived cells. Four different calcium-dependent release phenotypes were identified: two were specific for ACh or GABA, and two co-released two transmitters ACh and GABA but not Glu, or ACh and Glu but not GABA. We conclude that release mechanisms having different specificities are expressed by the cell lines studied, they become functional after loading the cells with the relevant transmitters. These observations will help the identification of proteins controlling the specificity of release, and provide an interesting model for pharmacological studies.