The complexity of large-scale neural networks can lead to poor understanding of their internal details. We show that this opaqueness provides an opportunity for adversaries to embed unintended functionalities into the network in the form of Trojan horses. Our novel framework hides the existence of a Trojan network with arbitrary desired functionality within a benign transport network. We prove theoretically that the Trojan network's detection is computationally infeasible and demonstrate empirically that the transport network does not compromise its disguise. Our paper exposes an important, previously unknown loophole that could potentially undermine the security and trustworthiness of machine learning.

TrojanNet: Embedding Hidden Trojan Horse Models in Neural Networks.

The deep learning model has a large number of parameters that need to be adjusted during the training process, so training a model that can perform well in object detection requires a large amount of valid data. And the accuracy and generalization ability of the model are guaranteed by both quantity and diversity of data. However, it’s a huge expense to get pictures and then annotate them, and we can hardly ever get all types of data that exist in reality. Apparently, data augmentation is an effective way to solve this problem. In our paper, the data set we use is composed of x-ray pictures come from subway security inspection. Based on the characteristics of x- ray pictures, we conducted experiments on a variety of augmentation methods, aimed at finding suitable data augmentation methods for this data set. After the experiment, it is concluded that the augmentation method that conforms to the actual form of the data set is more effective. Compared with the monotonous data augmentation, the hybrid way can be more effective, which can improve the accuracy and generalization ability of the model better. Meanwhile the complexity of backbone can also affect data augmentation, more complex backbone can be more sensitive to data augmentation.

Research on Data Augmentation for Object Detection Based on X- ray Security Inspection Picture

Radio spectrum resource is a scarce resource. How to use the limited spectrum resource to provide users with the maximum communication opportunities is worth further study. In fact, the traditional “command and control” mode of spectrum management is very inefficient in the utilization of spectrum, resulting in a great waste of frequency resources. At the same time, in radio management, it is necessary to set up databases of different services to store different monitoring information. Although the system plays a certain control ability, the radio monitoring records form an island of information, and the information of each service does not circulate, which is not conducive to integrated management. In this paper, based on the current radio spectrum resources in dynamic spectrum sharing method of resource sharing, the spectrum value of trust transfer problem, and the information management system each other information, collaborative management difficulties and other issues, put forward a kind of spectrum sharing method based on blockchain and by business combining intelligent contract database, realize data sharing solution.

Spectrum Resource Trading and Radio Management Data Sharing Based on Blockchain

Climate change issues societal operation, likely wanting considerable adaptation to deal with doing well altered weather patterns. Machine learning (ML) algorithms have progressed considerably, triggering breakthroughs in some other investigation sectors, along with only lately suggested as helping climate evaluation. Though a significant volume of isolated Earth System functions are analyzed with ML techniques, much more generic phone system to find out better the whole temperature unit hasn't happened. For instance, ML is able to aid remote identification, in which complex feedbacks make characterization tough from instantaneous equation analysis or perhaps possibly visualization of sizes plus Earth System design (ESM) diagnostics. Artificial intelligence (AI) may thus build on determined climate associates to provide enhanced alerts of approaching eco-friendly functions, which includes intense events. While ESM development is actually completely necessary, a parallel concentrate on utilizing ML and AI to determine as well as capitalize a great deal more on pre pre-existing simulations as well as info is suggested by us.

/pdf/artificial-intelligence-and-machine-learning-to-assist-12agtglrc5.pdf

Artificial Intelligence and Machine Learning to Assist Climate Change Monitoring

In terms of the current international situation, English is still the most widely used language interoperability many countries. Many people use English as a second language, and even some large international companies, who store their important documents in English, so research in the English language is needed, and one of the most significant research is the English text information filtering algorithm. As we all know, text information filtering is a kind of text information processing technology, including information classification, filtering and other technologies which is the link between text information and computer processing and filtering technology, and can be reprocessed according to the information content of the text. In the process of information processing, support vector machines (SVM) can solve many problems in the process. Generally, the problems such as high sparse dimension of text vector, high correlation between information features, and high sparsity among vectors always exist in text information filtering, while SVM can exactly solve the above problems. According to the feature of SVM for text problems, the center set region is established for positive sample clustering when filtering information, and it is difficult to judge the data and continue to use clustering decision when meeting later, so the accuracy of SVM algorithm in information filtering is improved.

Research on English Text Information Filtering Algorithm Based on SVM

Steganography is an art of obscuring data inside another quotidian file of similar or varying types. Hiding data has always been of significant importance to digital forensics. Previously, steganography has been combined with cryptography and neural networks separately. Whereas, this research combines steganography, cryptography with the neural networks all together to hide an image inside another container image of the larger or same size. Although the cryptographic technique used is quite simple, but is effective when convoluted with deep neural nets. Other steganography techniques involve hiding data efficiently, but in a uniform pattern which makes it less secure. This method targets both the challenges and make data hiding secure and non-uniform.

/pdf/hiding-data-in-images-using-cryptography-and-deep-neural-1f5957lw5v.pdf

Hiding Data in Images Using Cryptography and Deep Neural Network

/pdf/hiding-data-in-images-using-cryptography-and-deep-neural-1ba39fnav2.pdf

Sign language is a language used by people with hearing impairments to communicate using gestures Hand motion is one of the strategies utilized in gesture-based communication Through the years, many researchers have put forth their models for automatic sign language recognition The use of an RGB camera for recognition was a development, some of the models required signers to wear cyber gloves, etc, for ease in recognition Despite the fact that the utilization of gesture-based communication is mainstream among the hard of hearing individuals yet different groups do not attempt to learn it, this hinders communication and thus turns into a reason for the disconnection of certain groups in society There is a necessity of a framework that can encourage a method for communication between these two groups This paper provides a model that overcomes issues with using just an RGB camera The framework makes use of Kinect, an RGB camera and a depth sensor The model was developed using a convolutional neural network

Ashutosh Aggarwal

Papers

Hiding Data in Images Using Cryptography and Deep Neural Network

Hiding Data in Images Using Cryptography and Deep Neural Network

Sign Language Recognition Using Microsoft Kinect