Deep learning is at the heart of the current rise of artificial intelligence. In the field of Computer Vision, it has become the workhorse for applications ranging from self-driving cars to surveillance and security. Whereas deep neural networks have demonstrated phenomenal success (often beyond human capabilities) in solving complex problems, recent studies show that they are vulnerable to adversarial attacks in the form of subtle perturbations to inputs that lead a model to predict incorrect outputs. For images, such perturbations are often too small to be perceptible, yet they completely fool the deep learning models. Adversarial attacks pose a serious threat to the success of deep learning in practice. This fact has recently lead to a large influx of contributions in this direction. This article presents the first comprehensive survey on adversarial attacks on deep learning in Computer Vision. We review the works that design adversarial attacks, analyze the existence of such attacks and propose defenses against them. To emphasize that adversarial attacks are possible in practical conditions, we separately review the contributions that evaluate adversarial attacks in the real-world scenarios. Finally, drawing on the reviewed literature, we provide a broader outlook of this research direction.

Abstract-Viewpoint independent recognition of free-form objects and their segmentation in the presence of clutter and occlusions is a challenging task. We present a novel 3D model-based algorithm which performs this task automatically and efficiently. A 3D model of an object is automatically constructed offline from its multiple unordered range images (views). These views are converted into multidimensional table representations (which we refer to as tensors). Correspondences are automatically established between these views by simultaneously matching the tensors of a view with those of the remaining views using a hash table-based voting scheme. This results in a graph of relative transformations used to register the views before they are integrated into a seamless 3D model. These models and their tensor representations constitute the model library. During online recognition, a tensor from the scene is simultaneously matched with those in the library by casting votes. Similarity measures are calculated for the model tensors which receive the most votes. The model with the highest similarity is transformed to the scene and, if it aligns accurately with an object in the scene, that object is declared as recognized and is segmented. This process is repeated until the scene is completely segmented. Experiments were performed on real and synthetic data comprised of 55 models and 610 scenes and an overall recognition rate of 95 percent was achieved. Comparison with the spin images revealed that our algorithm is superior in terms of recognition rate and efficiency.

3D object recognition from local features is robust to occlusions and clutter. However, local features must be extracted from a small set of feature rich keypoints to avoid computational complexity and ambiguous features. We present an algorithm for the detection of such keypoints on 3D models and partial views of objects. The keypoints are highly repeatable between partial views of an object and its complete 3D model. We also propose a quality measure to rank the keypoints and select the best ones for extracting local features. Keypoints are identified at locations where a unique local 3D coordinate basis can be derived from the underlying surface in order to extract invariant features. We also propose an automatic scale selection technique for extracting multi-scale and scale invariant features to match objects at different unknown scales. Features are projected to a PCA subspace and matched to find correspondences between a database and query object. Each pair of matching features gives a transformation that aligns the query and database object. These transformations are clustered and the biggest cluster is used to identify the query object. Experiments on a public database revealed that the proposed quality measure relates correctly to the repeatability of keypoints and the multi-scale features have a recognition rate of over 95% for up to 80% occluded objects.

Abstract-We present a fully automatic face recognition algorithm and demonstrate its performance on the FRGC v2.0 data. Our algorithm is multimodal (2D and 3D) and performs hybrid (feature based and holistic) matching in order to achieve efficiency and robustness to facial expressions. The pose of a 3D face along with its texture is automatically corrected using a novel approach based on a single automatically detected point and the Hotelling transform. A novel 3D Spherical Face Representation (SFR) is used in conjunction with the Scale-Invariant Feature Transform (SIFT) descriptor to form a rejection classifier, which quickly eliminates a large number of candidate faces at an early stage for efficient recognition in case of large galleries. The remaining faces are then verified using a novel region-based matching approach, which is robust to facial expressions. This approach automatically segments the eyesforehead and the nose regions, which are relatively less sensitive to expressions and matches them separately using a modified Iterative Closest Point (ICP) algorithm. The results of all the matching engines are fused at the metric level to achieve higher accuracy. We use the FRGC benchmark to compare our results to other algorithms that used the same database. Our multimodal hybrid algorithm performed better than others by achieving 99.74 percent and 98.31 percent verification rates at a 0.001 false acceptance rate (FAR) and identification rates of 99.02 percent and 95.37 percent for probes with a neutral and a nonneutral expression, respectively.

Abstract