Defence methods for image adversarial attacks

In the previous post, we reviewed some well-known methods for black-box decision-based adversarial attacks where the adversary has no knowledge about the victim model except for its discrete hard-label predictions. Thus gradient-based methods become ineffective but simple random-walk-based methods such as the Boundary Attack can still represent a threat even under these particular settings. Now that we have introduced both white and black-box attacks under … Continue reading Defence methods for image adversarial attacks

Black-box decision-based attacks on images

In the previous post, we reviewed a series of black-box score-based adversarial attacks where the adversary has to estimate the gradient by querying the target model and retrieving the labels’ confidence score. In this post, we are going to explore the third category of black-box attacks, namely, black-box decision-based attacks. Under these settings, the only knowledge the attacker has about the model is only discrete … Continue reading Black-box decision-based attacks on images

Black-box score-based attacks on images

In the previous post, we reviewed a series of black-box transfer-based adversarial attacks where the adversary has to generate adversarial examples against a substitute model. In this post, we are going to explore the second category of black-box attacks, namely, black-box score-based attacks. Under this setting, it is not possible to access the white-box model’s gradients. The only knowledge about the attacked model is the … Continue reading Black-box score-based attacks on images

Black-box transfer-based attacks on images

In the previous post, we reviewed a series of white-box adversarial attacks where the adversary has full access to and knowledge of the victim model. In this post, we are going to explore the first category of black-box attacks, namely, black-box transfer-based attacks. Transfer-based attacks generate adversarial examples against a substitute model, possibly being as much similar as possible to the target model, which has … Continue reading Black-box transfer-based attacks on images

White-box adversarial attacks on images

In the first post, we introduced the concept of adversarial attacks and contextualized them in the case of images. In this post, we are going to explore the first category of attacks, namely, white-box attacks. Under this setting, the adversary has full access and knowledge of the model, that is, the architecture of the model, its parameters, gradients, and loss of respect to the input … Continue reading White-box adversarial attacks on images

Introduction to adversarial attacks on images

Nowadays, image classification deep learning models are always more present in our systems in order to create smarter applications or simply to replace human operators to automatically perform some repetitive tasks. Their increased utilization is due to their high accuracy such that recent models are now able to outperform human brains in many object classification tasks. However, despite their good generalization, deep neural networks are … Continue reading Introduction to adversarial attacks on images

MovieSearch: a smart movie search engine

MovieSearch is a content specific search engine with the aim to retrieve movie information given the contents of a user’s query. The search engine relies on the OkapiBM25 algorithm and takes into consideration the text present in the overview, the title, the names of the cast, and the production companies of each movie. The backend has been developed with the framework Django while the front-end … Continue reading MovieSearch: a smart movie search engine

Analysis of parallel version of PageRank algorithm

In this post, we are going to analyze a simplified parallel version of the famous algorithm PageRank, the algorithm used by Google Search to rank web pages in their search engine results. The code of the PageRank algorithm has been written in C++ exploiting the library OpenMP to parallelize the code. Finally, it has been tested over a different number of threads as well as different scheduling policies in order … Continue reading Analysis of parallel version of PageRank algorithm

The hacker, the photographer and the rival: Path Traversal

Artbit is an online social network where users can register for free and upload their own digital pictures in order to share them with the world and get popularity. To attract more visitors, the creators of the social network have launched a challenge on their platform which will reward the user who will upload the best picture. Any user can join the competition by simply … Continue reading The hacker, the photographer and the rival: Path Traversal

Playing Pacman with Multi-Agents Adversarial Search

In this post, we are going to design various artificial intelligence agents to play the classic version of Pacman, including ghosts and capsules. Pacman is a famous Atari game developed back in 1979 by a nine-person team and then released in 1980 by the former Japanese developer and publisher of arcade video games Namco. The great success the game had at the time, made it … Continue reading Playing Pacman with Multi-Agents Adversarial Search