Computer Vision publications
Permanent URI for this collection
Browse
Recent Submissions
Item Open Access ScanMix : Learning from Severe Label Noise via Semantic Clustering and Semi-Supervised Learning(Elsevier BV, 2023) Sachdeva, R.; Cordeiro, F.R.; Belagiannis, V.; Reid, I.; Carneiro, G.We propose a new training algorithm, ScanMix, that explores semantic clustering and semi-supervised learning (SSL) to allow superior robustness to severe label noise and competitive robustness to nonsevere label noise problems, in comparison to the state of the art (SOTA) methods. ScanMix is based on the expectation maximisation framework, where the E-step estimates the latent variable to cluster the training images based on their appearance and classification results, and the M-step optimises the SSL classification and learns effective feature representations via semantic clustering. We present a theoretical result that shows the correctness and convergence of ScanMix, and an empirical result that shows that ScanMix has SOTA results on CIFAR-10/-100 (with symmetric, asymmetric and semantic label noise), Red Mini-ImageNet (from the Controlled Noisy Web Labels), Clothing1M and WebVision. In all benchmarks with severe label noise, our results are competitive to the current SOTA.Item Metadata only The Road to Know-Where: An Object-and-Room Informed Sequential BERT for Indoor Vision-Language Navigation(IEEE, 2022) Qi, Y.; Pan, Z.; Hong, Y.; Yang, M.H.; Van Den Hengel, A.; Wu, Q.; IEEE/CVF International Conference on Computer Vision (ICCV) (10 Oct 2021 - 17 Oct 2021 : Montreal, QC, Canada (virtual online))Vision-and-Language Navigation (VLN) requires an agent to find a path to a remote location on the basis of natural-language instructions and a set of photo-realistic panoramas. Most existing methods take the words in the instructions and the discrete views of each panorama as the minimal unit of encoding. However, this requires a model to match different nouns (e.g., TV, table) against the same input view feature. In this work, we propose an object-informed sequential BERT to encode visual perceptions and linguistic instructions at the same fine-grained level, namely objects and words. Our sequential BERT also enables the visual-textual clues to be interpreted in light of the temporal context, which is crucial to multiround VLN tasks. Additionally, we enable the model to identify the relative direction (e.g., left/right/front/back) of each navigable location and the room type (e.g., bedroom, kitchen) of its current and final navigation goal, as such information is widely mentioned in instructions implying the desired next and final locations. We thus enable the model to know-where the objects lie in the images, and to know-where they stand in the scene. Extensive experiments demonstrate the effectiveness compared against several state-of-the-art methods on three indoor VLN tasks: REVERIE, NDH, and R2R. Project repository: https://github.com/YuankaiQi/ORISTItem Metadata only Satellite pose estimation with deep landmark regression and nonlinear pose refinement(IEEE, 2019) Chen, B.; Cao, J.; Parra Bustos, A.; Chin, T.; IEEE/CVF International Conference on Computer Vision Workshop (ICCVW) (27 Oct 2019 - 2 Nov 2019 : Seoul, South Korea)We propose an approach to estimate the 6DOF pose of a satellite, relative to a canonical pose, from a single image. Such a problem is crucial in many space proximity operations, such as docking, debris removal, and inter-spacecraft communications. Our approach combines machine learning and geometric optimisation, by predicting the coordinates of a set of landmarks in the input image, associating the landmarks to their corresponding 3D points on an a priori reconstructed 3D model, then solving for the object pose using non-linear optimisation. Our approach is not only novel for this specific pose estimation task, which helps to further open up a relatively new domain for machine learning and computer vision, but it also demonstrates superior accuracy and won the first place in the recent Kelvins Pose Estimation Challenge organised by the European Space Agency (ESA).Item Metadata only Deep learning for 2D scan matching and loop closure(IEEE, 2017) Li, J.; Zhan, H.; Chen, B.; Reid, I.; Lee, G.; IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (24 Sep 2017 - 28 Sep 2017 : Vancouver, Canada); Bicchi, A.; Okamura, A.Although 2D LiDAR based Simultaneous Localization and Mapping (SLAM) is a relatively mature topic nowadays, the loop closure problem remains challenging due to the lack of distinctive features in 2D LiDAR range scans. Existing research can be roughly divided into correlation based approaches e.g. scan-to-submap matching and feature based methods e.g. bag-of-words (BoW). In this paper, we solve loop closure detection and relative pose transformation using 2D LiDAR within an end-to-end Deep Learning framework. The algorithm is verified with simulation data and on an Unmanned Aerial Vehicle (UAV) flying in indoor environment. The loop detection ConvNet alone achieves an accuracy of 98.2% in loop closure detection. With a verification step using the scan matching ConvNet, the false positive rate drops to around 0.001%. The proposed approach processes 6000 pairs of raw LiDAR scans per second on a Nvidia GTX1080 GPU.Item Metadata only Data exfiltration: a review of external attack vectors and countermeasures(Elsevier, 2018) Ullah, F.; Edwards, M.; Ramdhany, R.; Chitchyan, R.; Babar, M.; Rashid, A.Context: One of the main targets of cyber-attacks is data exfiltration, which is the leakage of sensitive or private data to an unauthorized entity. Data exfiltration can be perpetrated by an outsider or an insider of an organization. Given the increasing number of data exfiltration incidents, a large number of data exfiltration countermeasures have been developed. These countermeasures aim to detect, prevent, or investigate exfiltration of sensitive or private data. With the growing interest in data exfiltration, it is important to review data exfiltration attack vectors and countermeasures to support future research in this field. Objective: This paper is aimed at identifying and critically analysing data exfiltration attack vectors and countermeasures for reporting the status of the art and determining gaps for future research. Method: We have followed a structured process for selecting 108 papers from seven publication databases. Thematic analysis method has been applied to analyse the extracted data from the reviewed papers. Results: We have developed a classification of (1) data exfiltration attack vectors used by external attackers and (2) the countermeasures in the face of external attacks. We have mapped the countermeasures to attack vectors. Furthermore, we have explored the applicability of various countermeasures for different states of data (i.e., in use, in transit, or at rest). Conclusion: This review has revealed that (a) most of the state of the art is focussed on preventive and detective countermeasures and significant research is required on developing investigative countermeasures that are equally important; (b) Several data exfiltration countermeasures are not able to respond in real-time, which specifies that research efforts need to be invested to enable them to respond in real-time (c) A number of data exfiltration countermeasures do not take privacy and ethical concerns into consideration, which may become an obstacle in their full adoption (d) Existing research is primarily focussed on protecting data in ‘in use’ state, therefore, future research needs to be directed towards securing data in ‘in rest’ and ‘in transit’ states (e) There is no standard or framework for evaluation of data exfiltration countermeasures. We assert the need for developing such an evaluation framework.Item Metadata only Three-Dimensional Endoscopy(Kluwer Academic, 2002) Thormaehlen, Thorsten; Broszio, Hellward; Meier, Peter N.; Falk Symposium (124th : 2001 : Hannover, Germany); School of Computer ScienceThis contribution presents a robust estimation technique for computation of camera motion and 3D reconstruction of a scene from an endoscopic image sequence. The technique offers new possibilities for more objective diagnoses and intervention planning. Every ordinary flexible monocular video endoscope can be used without modifications.