Category-Sensitive Hashing and Bloom Filter Based Descriptors for Online Keypoint Recognition

Bloom filters are widely used in many network applications but the high computation cost limits the system performance. In this paper, we introduce a new variation of Bloom filter named One-Hashing Bloom Filter (OHBF) to solve the problem. OHBF requires only one base hash function plus a few simple operations to implement a Bloom filter. While keeping nearly the same theoretical false positive ratio as an ideal Bloom filter, OHBF significantly reduces the hash computation overhead. We show that the false positive performance of a standard Bloom filter implementation strongly relies on the selection of hash functions, even if these hash functions are considered good. In contrast, OHBF presents consistently better performance with a proven mathematical foundation. OHBF is ideal for high throughput and low latency applications. As OHBF is a fundamental technique in Bloom filter theory, it can be applied to many other Bloom filter variations, such as Counting Bloom Filter and Space-Code Bloom Filter.

During the continual transitions from lab research to real-world applications of vision-based algorithms, there are significant challenges, e.g. the robustness to adapt to complex environments and the high demands of multi-task, multi-modal learning models. More specifically on visual simultaneous localization and mapping (visual SLAM) for mobile robots, two notable limitations are: (i) the drift issue during pose estimation, especially in dynamic environments, makes the positioning system unstable; (ii) the additional loop detection takes many computational resources for image retrieval and global feature geometry checking. This thesis explores the approaches of visual localization on unmanned ground vehicles by solving those limitations. The conventional feature extraction and matching pipeline for vision-based tasks involve three steps: (i) local feature detection and description, (ii) feature matching, and (iii) outlier rejection. The first contribution of the thesis is adding a feature selection and anticipation stage to reduce tracking drift. We explore the noise model of image features to select a subset of all the observed image features with the best ”contribution” during data association and pose estimation across multiple frames. Conventional SLAM algorithms take a strong assumption of scene rigidity, which limits the application under challenging environments. The second part of the thesis addresses the tough issue in dynamic environments with moving objects. We presented GMC, namely the motion clustering approach, a lightweight dynamic object filtering method. It can distinguish moving objects from static landmarks. Based on the theory of motion coherence within a particular image area, GMC could segment dynamic objects in 3D space. We can provide an efficient and robust correspondence algorithm that can extract dynamic objects from a static background with the method. In this way, we propose a dynamic SLAM system that is real-time and free from expensive GPU processors. In contrast to GMC, the thesis’s third part turns to an object-aware learning-based model for more general dynamic scenarios. We use object detection and tracking as points, lines, planes, etc. We utilize semantic information and extract sparse image features simultaneously to keep track of dynamic objects. The static background and different dynamic objects are jointly optimized in a newly developed bundle adjustment sliding window. The estimated 3D bounding boxes can provide more robust camera tracking and better scene understanding, and better map merging. The fourth part of the thesis leverages the emerging feature learning framework. It proposes a unified self-supervised model called LGDNet to generate both global and local image feature descriptors end-to-end. Global feature descriptors embed the whole image into a compact representation, leading to easier scene comparison. On the other hand, local features focus more on the local region similarities of some salient parts for structure from motion. Our proposed method can directly extract features together with descriptors that encode both local maximum responses and global context information, avoiding duplicate calculations based on different feature extraction criteria.

Probabilistic techniques for verification of finite-state transition systems offer huge memory savings over deterministic techniques. The two leading probabilistic schemes are hash compaction and the bitstate method, which stores states in a Bloom filter. Bloom filters have been criticized for being slow, inaccurate, and memory-inefficient, but in this paper, we show how to obtain Bloom filters that are simultaneously fast, accurate, memory-efficient, scalable, and flexible. The idea is that we can introduce large dependences among the hash functions of a Bloom filter with almost no observable effect on accuracy, and because computation of independent hash functions was the dominant computational cost of accurate Bloom filters and model checkers based on them, our savings are tremendous. We present a mathematical analysis of Bloom filters in verification in unprecedented detail, which enables us to give a fresh comparison between hash compaction and Bloom filters. Finally, we validate our work and analyses with extensive testing using 3SPIN, a model checker we developed by extending SPIN.

deepBF: Malicious URL detection using learned Bloom Filter and evolutionary deep learning

A Bloom filter is a simple space-efficient randomized data structure for representing a set in order to support network and database query systems. Although Bloom filters were invented in the 1970's and have been heavily used in database applications, they have only recently received widespread attention in the networking literature. Software applications frequently fail to identify so many signatures through comparisons at very high speeds. We introduce a low power bloom filter architectures which is space and power effective in hardware platforms. Instead of working on programming phase or technology, our work concentrated on look up techniques of Bloom filters. We have chosen better hash function to be utilized in hardware applications of Bloom filter by FPGA implementation. Power analysis shows that decrement in number of hash functions per stage results in power gain.

Bloom filters are a type of data structure that is used to test whether or not an element is a member of a set. They are known for being space-efficient and are commonly employed in various applications, such as network routers, web browsers, and databases. These filters work by allowing a fixed probability of incorrectly identifying an element as being a member of the set, known as the false positive rate (FPR). However, traditional bloom filters suffer from a high FPR and extensive memory usage, which can lead to incorrect query results and a slow performance. Thus, this study indicates that a content-based strategy could be a practical solution for these challenges. Specifically, our approach requires less bloom filter storage, consequently decreasing the probability of false positives. The effectiveness of several hash functions on our strategy’s performance was also evaluated. Experimental evaluations demonstrated that the proposed strategy could potentially decrease false positives by a substantial margin of up to 79.83%. The use of size-based content bits significantly contributes to the decrease in the number of false positives as well. However, as the volume of content bits rises, the impact on time is not considerably noticeable. Moreover, the evidence suggests that the application of a singular approach leads to a more than 50% decrease in false positives.

Bloom filters are widely used in networking and computing to accelerate membership checking. In many applications filters store sensitive data, so their privacy is of primary concern. At first glance, it seems that extracting the set of elements inserted from the filter would not be possible, because in Bloom filters elements are mapped to positions using hash functions. However, previous works have shown that for the Bloom filter, it may be possible to identify few of the elements inserted in the filter. In this work, we consider the case of counting Bloom filters (CBFs) and show that in some cases, the entire set of elements used to create the filter can be extracted from the filter. This poses serious privacy and security concerns when an attacker can get access to the filter contents. In this article, an algorithm to extract the elements inserted from the filter is presented and analyzed theoretically; then, the feasibility of the CBF inversion is shown by simulation. A case study is presented in detail to illustrate that in practical applications, these conditions can be met by using additional restrictions that are implicit in the nature of the application itself.

Hashing is an important function in many applications such as hash tables, caches and Bloom filters. In past, genetic programming was applied to evolve application-specific as well as general-purpose hash functions, where the main design target was the quality of hashing. As hash functions are frequently called in various time-critical applications, it is important to optimize their implementation with respect to the execution time. In this paper, linear genetic programming is combined with NSGA-II algorithm in order to obtain general-purpose, ultra-fast and high-quality hash functions. Evolved hash functions show highly competitive quality of hashing, but significantly reduced execution time in comparison with the state of the art hash functions available in literature.

In computing, duplicate data detection refers to identifying duplicate copies of repeating data. Identifying duplicate data items in streamed data and eliminating them before storing, is a complex job. This paper proposes a novel data structure for duplicate detection using a variant of stable Bloom filter named as FingerPrint Stable Bloom Filter (FP-SBF). The proposed approach uses counting Bloom filter with fingerprint bits along with an optimization mechanism for duplicate detection. FP-SBF uses d-left hashing which reduces the computational time and decreases the false positives as well as false negatives. FP-SBF can process unbounded data in single pass, using k hash functions, and successfully differentiate between duplicate and distinct elements in O(k+1) time, independent of the size of incoming data. The performance of FP-SBF has been compared with various Bloom Filters used for stream data duplication detection and it has been theoretically and experimentally proved that the proposed approach efficiently detects the duplicates in streaming data with less memory requirements.

To solve the low accuracy of image feature matching in horticultural robot visual navigation, an innovative and effective image feature matching algorithm was proposed combining the improved Oriented FAST and Rotated BRIEF (ORB) and Lucas–Kanade (LK) optical flow algorithm. First, image feature points were extracted according to the adaptive threshold calculated using the Michelson contrast. Then, the extracted feature points were uniformed by the quadtree structure, which can reduce the calculated amount of feature matching, and the uniform ORB feature points were roughly matched to estimate the position of the feature points in the matched image using the improved LK optical flow. Finally, the Hamming distance between rough matching points was calculated for precise matching. Feature extraction and matching experiments were performed in four typical scenes: normal light, low light, high texture, and low texture. Compared with the traditional algorithm, the uniformity and accuracy of the feature points extracted by the proposed algorithm were enhanced by 0.22 and 50.47%, respectively. Meanwhile, the results revealed that the matching accuracy of the proposed algorithm increased by 14.59%, whereas the matching time and total time decreased by 39.18% and 44.79%, respectively. The proposed algorithm shows great potential for application in the visual simultaneous localization and mapping (V-SLAM) of horticultural robots to achieve higher accuracy of real-time positioning and map construction.

The current visual simultaneous localization and mapping (SLAM) systems require the use of matched feature point pairs to estimate camera pose and construct environmental maps. Therefore, they suffer from poor performance of the visual feature matchers. To address this problem, a visual SLAM using deep feature matcher is proposed, which is mainly composed of three parallel threads: Visual Odometry, Backend Optimizer and LoopClosing. In the Visual Odometry, the deep feature extractor with convolutional neural networks is utilized for extracting feature points in each image frame. Then, the deep feature matcher is used for obtaining the corresponding feature landmark pairs. Afterwards, a fusion method based on the last and the reference frame is proposed for camera pose estimation. The Backend Optimizer is designed to execute local bundle adjustment for a part of camera poses and landmarks (map points). While LoopClosing, consisting of a lightweight deep loop closure detector and the same matcher as the one used in Visual Odometry is utilized for loop correction based on pose graph. The proposed system has been tested extensively on most of benchmark KITTI odometry dataset. The experimental results show that our system yields better performance than the existing visual SLAM systems. It can not only run in real-time at a speed of 0.08 s per frame, but also reduce estimation error by at least 0.1 m.

The indoor Visual Simultaneous Localization And Mapping (V-SLAM) dataset with various acquisition modalities has been created to evaluate the impact of acquisition modalities on the Visual SLAM algorithm’s accuracy. The dataset contains different sequences acquired with different modalities, including RGB, IR, and depth images in passive stereo and active stereo modes. Each sequence is associated with a reference trajectory constructed with an Structure From Motion (SFM) and Multi View Stereo (MVS) library for comparison. Data were collected using an intrinsically calibrated Intel RealSense D435i camera. The RGB/IR and depth data are spatially aligned, and the stereo images are rectified. The dataset includes various areas, some with low brightness, with changes in brightness, wide, narrow and texture.

We present segmentation and tracking of deformable objects using non-linear on-line learning of high-level shape information in the form of a level set function.The emphasis is for successful tracking of objects that undergo smooth arbitrary deformations, but without the a priori learning of shape constraints.The high-level shape information is learnt on-line by defining a memory of object samples in a high-dimensional shape space.These shape samples are then used as weights via a locally defined shape space kernel function to define a template against which potential future shapes of the tracked object can be compared.Results for the successful tracking of a range of deformable motions are presented.

Distilled representation using patch-based local-to-global similarity strategy for visual place recognition

Future Internet requires re-thinking of network infrastructure towards the balance between computing capacities and energy sustainable techniques. As one of computing intensive components, Bloom filters are widely used for network packet processing. In this paper, an energy efficient self-adaptive Bloom filter, EABF, is devoted to a balance of power and performance especially for high performance networks. The basic idea is to give the Bloom Filter the capability to adjust the number of active hash functions according to the current workload automatically. This adaption depends on its control policies. Three policies are presented and compared. We also give the method to implement EABF in hardware for higher performance. It is presented in a two-stage platform based on FPGA where Stage 1 is always active and Stage 2, a secondary stage, is only active when necessary. The platform can also be extended to multi-stages. A control circuit is designed for flexibly changing working stage and reducing both dynamic and static power consumption. Analysis and experiments show that our dynamic two-stage EABF can achieve almost the best power savings as that of the fixed schemes; unlike the fixed schemes that might have much longer latency, EABF maintains nearly 1 clock cycle latency as that of a regular Bloom filter.