Few-shot detection-based counters estimate the number of instances in the image specified only by a few test-time exemplars. A common approach to localize objects across multiple sizes is to merge backbone features of different resolutions. Furthermore, to enable small object detection in densely populated regions, the input image is commonly upsampled and tiling is applied to cope with the increased computational and memory requirements. Because of these ad-hoc solutions, existing counters struggle with images containing diverse-sized objects and densely populated regions of small objects. We propose GeCo2, an end-to-end few-shot counting and detection method that explicitly addresses the object scale issues. A new dense query representation gradually aggregates exemplar-specific feature information across scales that leads to high-resolution dense queries that enable detection of large as well as small objects. GeCo2 surpasses state-of-the-art few-shot counters in counting as well as detection accuracy by ~10% while running ~3x faster at smaller GPU memory footprint.

Memory-based trackers are video object segmentation methods that form the target model by concatenating recently tracked frames into a memory buffer and localize the target by attending the current image to the buffered frames. While already achieving top performance on many benchmarks, it was the recent release of SAM2 that placed memory-based trackers into focus of the visual object tracking community. Nevertheless, modern trackers still struggle in the presence of distractors. We argue that a more sophisticated memory model is required, and propose a new distractor-aware memory model for SAM2 and an introspection-based update strategy that jointly addresses the segmentation accuracy as well as tracking robustness. The resulting tracker is denoted as DAM4SAM. We also propose a new distractor-distilled DiDi dataset to study the distractor problem better. DAM4SAM outperforms SAM2.1 and related SAM memory extensions on seven benchmarks and sets a solid new state-of-the-art on six of them. The code and the new dataset are available on https://github.com/jovanavidenovic/DAM4SAM.

Low-shot counters estimate the number of objects corresponding to a selected category, based on only few or no exemplars annotated in the image. The current state-of-the-art estimates the total counts as the sum over the object location density map, but does not provide individual object locations and sizes, which are crucial for many applications. This is addressed by detection-based counters, which, however fall behind in the total count accuracy. Furthermore, both approaches tend to overestimate the counts in the presence of other object classes due to many false positives. We propose DAVE, a low-shot counter based on a detect-and-verify paradigm, that avoids the aforementioned issues by first generating a high-recall detection set and then verifying the detections to identify and remove the out-liers. This jointly increases the recall and precision, leading to accurate counts. DAVE outperforms the top density-based counters by ~20% in the total count MAE, it outper-forms the most recent detection-based counter by ~20% in detection quality and sets a new state-of-the-art in zero-shot as well as text-prompt-based counting. The code and models are available on GitHub.

Recently a growing number of works focus on machine defect detection from anomalous audio patterns. The datasets for the machine audio domain are scarce and recent methods that perform well on benchmarks such as DCASE2020 Task 2, rely on auxiliary information such as annotated data from other training classes in the domain to extract information that can be used in deep-learning classification-based anomaly detection approaches. However, in practical scenarios, annotated data from the same domain may not be readily available so annotation-free methods that can learn appropriate audio representations from unannotated data are needed. We propose AudDSR, a simulation-based anomaly detection method that learns to detect anomalies without additional annotated data and instead focuses on a discrete feature space sampling method for an anomaly simulation process. AudDSR outperforms competing methods that do not rely on annotated data on the DCASE2020 anomalous sound detection benchmark and even matches the performance of some methods that utilize additional annotation information.

Low-shot object counters estimate the number of objects in an image using few or no annotated exemplars. Objects are localized by matching them to prototypes, which are constructed by unsupervised image-wide object appearance aggregation. Due to potentially diverse object appearances, the existing approaches often lead to overgeneralization and false positive detections. Furthermore, the best-performing methods train object localization by a surrogate loss, that predicts a unit Gaussian at each object center. This loss is sensitive to annotation error, hyperparameters and does not directly optimize the detection task, leading to suboptimal counts. We introduce GeCo, a novel low-shot counter that achieves accurate object detection, segmentation, and count estimation in a unified architecture. GeCo robustly generalizes the prototypes across objects appearances through a novel dense object query formulation. In addition, a novel counting loss is proposed, that directly optimizes the detection task and avoids the issues of the standard surrogate loss. GeCo surpasses the leading few-shot detection-based counters by $\sim$25\% in the total count MAE, achieves superior detection accuracy and sets a new solid state-of-the-art result across all low-shot counting setups. The code will be available on GitHub.

RGB-based surface anomaly detection methods have advanced significantly. However, certain surface anomalies remain practically invisible in RGB alone, necessitating the incorporation of 3D information. Existing approaches that employ point-cloud backbones suffer from suboptimal representations and reduced applicability due to slow processing. Re-training RGB backbones, designed for faster dense input processing, on industrial depth datasets is hindered by the limited availability of sufficiently large datasets. We make several contributions to address these challenges. (i) We propose a novel Depth-Aware Discrete Autoencoder (DADA) architecture, that enables learning a general discrete latent space that jointly models RGB and 3D data for 3D surface anomaly detection. (ii) We tackle the lack of diverse industrial depth datasets by introducing a simulation process for learning informative depth features in the depth encoder. (iii) We propose a new surface anomaly detection method 3DSR, which outperforms all existing state-of-theart on the challenging MVTec3D anomaly detection benchmark, both in terms of accuracy and processing speed. The experimental results validate the effectiveness and efficiency of our approach, highlighting the potential of utilizing depth information for improved surface anomaly detection. Code is available at: https://github.com/VitjanZ/3DSR

Abstract Performance of modern trackers degrades substantially on transparent objects compared to opaque objects. This is largely due to two distinct reasons. Transparent objects are unique in that their appearance is directly affected by the background. Furthermore, transparent object scenes often contain many visually similar objects (distractors), which often lead to tracking failure. However, development of modern tracking architectures requires large training sets, which do not exist in transparent object tracking. We present two contributions addressing the aforementioned issues. We propose the first transparent object trackingtraining datasetTrans2k that consists of over 2k sequences with 104,343 images overall, annotated by bounding boxes and segmentation masks. Standard trackers trained on this dataset consistently improve by up to 16%. Our second contribution is a new distractor-aware transparent object tracker (DiTra) that treats localization accuracy and target identification as separate tasks and implements them by a novel architecture. DiTra sets a new state-of-the-art in transparent object tracking and generalizes well to opaque objects.

We consider low-shot counting of arbitrary semantic categories in the image using only few annotated exemplars (few-shot) or no exemplars (no-shot). The standard few-shot pipeline follows extraction of appearance queries from exemplars and matching them with image features to infer the object counts. Existing methods extract queries by feature pooling which neglects the shape information (e.g., size and aspect) and leads to a reduced object localization accuracy and count estimates.We propose a Low-shot Object Counting network with iterative prototype Adaptation (LOCA). Our main contribution is the new object prototype extraction module, which iteratively fuses the exemplar shape and appearance information with image features. The module is easily adapted to zero-shot scenarios, enabling LOCA to cover the entire spectrum of low-shot counting problems. LOCA outperforms all recent state-of-the-art methods on FSC147 benchmark by 20-30% in RMSE on one-shot and few-shot and achieves state-of-the-art on zero-shot scenarios, while demonstrating better generalization capabilities. The code and models are available here: https://github.com/djukicn/loca.

The progress in maritime obstacle detection is hindered by the lack of a diverse dataset that adequately captures the complexity of general maritime environments. We present the first maritime panoptic obstacle detection benchmark LaRS, featuring scenes from Lakes, Rivers and Seas. Our major contribution is the new dataset, which boasts the largest diversity in recording locations, scene types, obstacle classes, and acquisition conditions among the related datasets. LaRS is composed of over 4000 per-pixel labeled key frames with nine preceding frames to allow utilization of the temporal texture, amounting to over 40k frames. Each key frame is annotated with 8 thing, 3 stuff classes and 19 global scene attributes. We report the results of 27 semantic and panoptic segmentation methods, along with several performance insights and future research directions. To enable objective evaluation, we have implemented an online evaluation server. The LaRS dataset, evaluation toolkit and benchmark are publicly available at: https://lojzezust.github.io/lars-dataset

Robust maritime obstacle detection is essential for fully autonomous unmanned surface vehicles (USVs). The currently widely adopted segmentation-based obstacle detection methods are prone to misclassification of object reflections and sun glitter as obstacles, producing many false positive detections, effectively rendering the methods impractical for USV navigation. However, water-turbulence-induced temporal appearance changes on object reflections are very distinctive from the appearance dynamics of true objects. We harness this property to design WaSR-T, a novel maritime obstacle detection network, that extracts the temporal context from a sequence of recent frames to reduce ambiguity. By learning the local temporal characteristics of object reflection on the water surface, WaSR-T substantially improves obstacle detection accuracy in the presence of reflections and glitter. Compared with existing single-frame methods, WaSR-T reduces the number of false positive detections by 41% overall and by over 53% within the danger zone of the boat, while preserving a high recall, and achieving new state-of-the-art performance on the challenging MODS maritime obstacle detection benchmark. The code, pretrained models and extended datasets are available at: https://github.com/lojzezust/WaSR-T

Coastal water autonomous boats rely on robust perception methods for obstacle detection and timely collision avoidance. The current state-of-the-art is based on deep segmentation networks trained on large datasets. Per-pixel ground truth labeling of such datasets, however, is labor-intensive and expensive. We observe that far less information is required for practical obstacle avoidance - the location of water edge on static obstacles like shore and approximate location and bounds of dynamic obstacles in the water is sufficient to plan a reaction. We propose a new scaffolding learning regime (SLR) that allows training obstacle detection segmentation networks only from such weak annotations, thus significantly reducing the cost of ground-truth labeling. Experiments show that maritime obstacle segmentation networks trained using SLR substantially outperform the same networks trained with dense ground truth labels, despite a significant reduction in labelling effort. Thus accuracy is not sacrificed for labelling simplicity but is in fact improved, which is a remarkable result.

Template-based discriminative trackers are currently the dominant tracking paradigm due to their robustness, but are restricted to bounding box tracking and a limited range of transformation models, which reduces their localization accuracy. We propose a discriminative single-shot segmentation tracker – D3S$_2$, which narrows the gap between visual object tracking and video object segmentation. A single-shot network applies two target models with complementary geometric properties, one invariant to a broad range of transformations, including non-rigid deformations, the other assuming a rigid object to simultaneously achieve robust online target segmentation. The overall tracking reliability is further increased by decoupling the object and feature scale estimation. Without per-dataset finetuning, and trained only for segmentation as the primary output, D3S$_2$outperforms all published trackers on the recent short-term tracking benchmark VOT2020 and performs very close to the state-of-the-art trackers on the GOT-10k, TrackingNet, OTB100 and LaSoT. D3S$_2$outperforms the leading segmentation tracker SiamMask on video object segmentation benchmarks and performs on par with top video object segmentation algorithms.

Obstacle detection using semantic segmentation has become an established approach in autonomous vehicles. However, existing segmentation methods, primarily developed for ground vehicles, are inadequate in an aquatic environment as they produce many false positive (FP) detections in the presence of water reflections and wakes. We propose a novel deep encoder-decoder architecture, a water segmentation and refinement (WaSR) network, specifically designed for the marine environment to address these issues. A deep encoder based on ResNet101 with atrous convolutions enables the extraction of rich visual features, while a novel decoder gradually fuses them with inertial information from the inertial measurement unit (IMU). The inertial information greatly improves the segmentation accuracy of the water component in the presence of visual ambiguities, such as fog on the horizon. Furthermore, a novel loss function for semantic separation is proposed to enforce the separation of different semantic components to increase the robustness of the segmentation. We investigate different loss variants and observe a significant reduction in FPs and an increase in true positives (TPs). Experimental results show that WaSR outperforms the current state of the art by approximately 4% in F1 score on a challenging unmanned surface vehicle dataset. WaSR shows remarkable generalization capabilities and outperforms the state of the art by over 24% in F1 score on a strict domain generalization experiment.

Small-sized unmanned surface vehicles (USV) are coastal water devices with a broad range of applications such as environmental control and surveillance. A crucial capability for autonomous operation is obstacle detection for timely reaction and collision avoidance, which has been recently explored in the context of camera-based visual scene interpretation. Owing to curated datasets, substantial advances in scene interpretation have been made in a related field of unmanned ground vehicles. However, the current maritime datasets do not adequately capture the complexity of real-world USV scenes and the evaluation protocols are not standardised, which makes cross-paper comparison of different methods difficult and hinders the progress. To address these issues, we introduce a new obstacle detection benchmark MODS, which considers two major perception tasks: maritime object detection and the more general maritime obstacle segmentation. We present a new diverse maritime evaluation dataset containing approximately 81k stereo images synchronized with an on-board IMU, with over 60k objects annotated. We propose a new obstacle segmentation performance evaluation protocol that reflects the detection accuracy in a way meaningful for practical USV navigation. Nineteen recent state-of-the-art object detection and obstacle segmentation methods are evaluated using the proposed protocol, creating a benchmark to facilitate development of the field. The proposed dataset, as well as evaluation routines, are made publicly available at vicos.si/resources.

Visual surface anomaly detection aims to detect local image regions that significantly deviate from normal appearance. Recent surface anomaly detection methods rely on generative models to accurately reconstruct the normal areas and to fail on anomalies. These methods are trained only on anomaly-free images, and often require hand-crafted post-processing steps to localize the anomalies, which prohibits optimizing the feature extraction for maximal detection capability. In addition to reconstructive approach, we cast surface anomaly detection primarily as a discriminative problem and propose a discriminatively trained reconstruction anomaly embedding model (DRÆM). The proposed method learns a joint representation of an anomalous image and its anomaly-free reconstruction, while simultaneously learning a decision boundary between normal and anomalous examples. The method enables direct anomaly localization without the need for additional complicated post-processing of the network output and can be trained using simple and general anomaly simulations. On the challenging MVTec anomaly detection dataset, DRÆM outperforms the current state-of-the-art unsupervised methods by a large margin and even de-livers detection performance close to the fully-supervised methods on the widely used DAGM surface-defect detection dataset, while substantially outperforming them in localization accuracy. Code at github.com/VitjanZ/DRAEM.

A long-term visual object tracking performance evaluation methodology and a benchmark are proposed. Performance measures are designed by following a long-term tracking definition to maximize the analysis probing strength. The new measures outperform existing ones in interpretation potential and in better distinguishing between different tracking behaviors. We show that these measures generalize the short-term performance measures, thus linking the two tracking problems. Furthermore, the new measures are highly robust to temporal annotation sparsity and allow annotation of sequences hundreds of times longer than in the current datasets without increasing manual annotation labor. A new challenging dataset of carefully selected sequences with many target disappearances is proposed. A new tracking taxonomy is proposed to position trackers on the short-term/long-term spectrum. The benchmark contains an extensive evaluation of the largest number of long-term trackers and comparison to state-of-the-art short-term trackers. We analyze the influence of tracking architecture implementations to long-term performance and explore various redetection strategies as well as the influence of visual model update strategies to long-term tracking drift. The methodology is integrated in the VOT toolkit to automate experimental analysis and benchmarking and to facilitate the future development of long-term trackers.

Template-based discriminative trackers are currently the dominant tracking paradigm due to their robustness, but are restricted to bounding box tracking and a limited range of transformation models, which reduces their localization accuracy. We propose a discriminative single-shot segmentation tracker - D3S, which narrows the gap between visual object tracking and video object segmentation. A single-shot network applies two target models with complementary geometric properties, one invariant to a broad range of transformations, including non-rigid deformations, the other assuming a rigid object to simultaneously achieve high robustness and online target segmentation. Without per-dataset finetuning and trained only for segmentation as the primary output, D3S outperforms all trackers on VOT2016, VOT2018 and GOT-10k benchmarks and performs close to the state-of-the-art trackers on the TrackingNet. D3S outperforms the leading segmentation tracker SiamMask on video segmentation benchmark and performs on par with top video object segmentation algorithms, while running an order of magnitude faster, close to real-time.

The best RGBD trackers provide high accuracy but are slow to run. On the other hand, the best RGB trackers are fast but clearly inferior on the RGBD datasets. In this work, we propose a deep depth-aware long-term tracker that achieves state-of-the-art RGBD tracking performance and is fast to run. We reformulate deep discriminative correlation filter (DCF) to embed the depth information into deep features. Moreover, the same depth-aware correlation filter is used for target redetection. Comprehensive evaluations show that the proposed tracker achieves state-of-the-art performance on the Princeton RGBD, STC, and the newly-released CDTB benchmarks and runs 20 fps.

Obstacle detection by semantic segmentation shows a great promise for autonomous navigation in unmanned surface vehicles (USV). However, existing methods suffer from poor estimation of the water edge in presence of visual ambiguities, poor detection of small obstacles and high false-positive rate on water reflections and wakes. We propose a new deep encoder-decoder architecture, a water-obstacle separation and refinement network (WaSR), to address these issues. Detection and water edge accuracy are improved by a novel decoder that gradually fuses inertial information from inertial measurement unit (IMU) with the visual features from the encoder. In addition, a novel loss function is designed to increase the separation between water and obstacle features early on in the network. Subsequently, the capacity of the remaining layers in the decoder is better utilised, leading to a significant reduction in false positives and increased true positives. Experimental results show that WaSR outperforms the current state-of-the-art by a large margin, yielding a 14% increase in F-measure over the second-best method.

Standard RGB-D trackers treat the target as a 2D structure, which makes modelling appearance changes related even to out-of-plane rotation challenging. This limitation is addressed by the proposed long-term RGB-D tracker called OTR - Object Tracking by Reconstruction. OTR performs online 3D target reconstruction to facilitate robust learning of a set of view-specific discriminative correlation filters (DCFs). The 3D reconstruction supports two performance- enhancing features: (i) generation of an accurate spatial support for constrained DCF learning from its 2D projection and (ii) point-cloud based estimation of 3D pose change for selection and storage of view-specific DCFs which robustly localize the target after out-of-view rotation or heavy occlusion. Extensive evaluation on the Princeton RGB-D tracking and STC Benchmarks shows OTR outperforms the state-of-the-art by a large margin.

We propose a new color-and-depth general visual object tracking benchmark (CDTB). CDTB is recorded by several passive and active RGB-D setups and contains indoor as well as outdoor sequences acquired in direct sunlight. The CDTB dataset is the largest and most diverse dataset in RGB-D tracking, with an order of magnitude larger number of frames than related datasets. The sequences have been carefully recorded to contain significant object pose change, clutter, occlusion, and periods of long-term target absence to enable tracker evaluation under realistic conditions. Sequences are per-frame annotated with 13 visual attributes for detailed analysis. Experiments with RGB and RGB-D trackers show that CDTB is more challenging than previous datasets. State-of-the-art RGB trackers outperform the recent RGB-D trackers, indicating a large gap between the two fields, which has not been previously detected by the prior benchmarks. Based on the results of the analysis we point out opportunities for future research in RGB-D tracker design.

The progress of obstacle detection via semantic segmentation on unmanned surface vehicles (USVs) has been significantly lagging behind the developments in the related field of autonomous cars. The reason is the lack of large curated training datasets from USV domain required for development of data-hungry deep CNNs. This paper addresses this issue by presenting MaSTr1325, a marine semantic segmentation training dataset tailored for development of obstacle detection methods in small-sized coastal USVs. The dataset contains 1325 diverse images captured over a two year span with a real USV, covering a range of realistic conditions encountered in a coastal surveillance task. The images are per-pixel semantically labeled. The dataset exceeds previous attempts in this domain in size, scene complexity and domain realism. In addition, a dataset augmentation protocol is proposed to address slight appearance differences of the images in the training set and those in deployment. The accompanying experimental evaluation provides a detailed analysis of popular deep architectures, annotation accuracy and influence of the training set size. MaSTr1325 will be released to research community to facilitate progress in obstacle detection for USVs.

Classical deep convolutional networks increase receptive field size by either gradual resolution reduction or application of hand-crafted dilated convolutions to prevent increase in the number of parameters. In this paper we propose a novel displaced aggregation unit (DAU) that does not require hand-crafting. In contrast to classical filters with units (pixels) placed on a fixed regular grid, the displacement of the DAUs are learned, which enables filters to spatially-adapt their receptive field to a given problem. We extensively demonstrate the strength of DAUs on a classification and semantic segmentation tasks. Compared to ConvNets with regular filter, ConvNets with DAUs achieve comparable performance at faster convergence and up to 3-times reduction in parameters. Furthermore, DAUs allow us to study deep networks from novel perspectives. We study spatial distributions of DAU filters and analyze the number of parameters allocated for spatial coverage in a filter.

We propose a stereo-based obstacle detection approach for unmanned surface vehicles. Obstacle detection is cast as a scene semantic segmentation problem in which pixels are assigned a probability of belonging to water or non-water regions. We extend a single-view model to a stereo system by adding a constraint which prefers consistent class labels assignment to pixels in the left and right camera images corresponding to the same parts of a 3D scene. Our approach jointly fits a semantic model to both images, leading to an improved class-label posterior map from which obstacles and water edge are extracted. In overall F-measure, our approach outperforms the current state-of-the-art monocular approach by 0.495, a monocular CNN by 0.798 and their stereo extensions by 0.059 and 0.515, respectively on the task of obstacle detection while running real-time on a single CPU.

Deformable parts models show a great potential in tracking by principally addressing nonrigid object deformations and self occlusions, but according to recent benchmarks, they often lag behind the holistic approaches. The reason is that potentially large number of degrees of freedom have to be estimated for object localization and simplifications of the constellation topology are often assumed to make the inference tractable. We present a new formulation of the constellation model with correlation filters that treats the geometric and visual constraints within a single convex cost function and derive a highly efficient optimization for maximum a posteriori inference of a fully connected constellation. We propose a tracker that models the object at two levels of detail. The coarse level corresponds a root correlation filter and a novel color model for approximate object localization, while the mid-level representation is composed of the new deformable constellation of correlation filters that refine the object location. The resulting tracker is rigorously analyzed on a highly challenging OTB, VOT2014, and VOT2015 benchmarks, exhibits a state-of-the-art performance and runs in real-time.

Short-term tracking is an open and challenging problem for which discriminative correlation filters (DCF) have shown excellent performance.We introduce the channel and spatial reliability concepts to DCF tracking and provide a novel learning algorithm for its efficient and seamless integration in the filter update and the tracking process. The spatial reliability map adjusts the filter support to the part of the object suitable for tracking. This both allows to enlarge the search region and improves tracking of non-rectangular objects. Reliability scores reflect channel-wise quality of the learned filters and are used as feature weighting coefficients in localization. Experimentally, with only two simple standard features, HoGs and Colornames, the novel CSRDCF method - DCF with Channel and Spatial Reliability - achieves state-of-the-art results on VOT 2016, VOT 2015 and OTB100. The CSR-DCF runs in real-time on a CPU.

Object-to-camera motion produces a variety of apparent motion patterns that significantly affect performance of short-term visual trackers. Despite being crucial for designing robust trackers, their influence is poorly explored in standard benchmarks due to weakly defined, biased and overlapping attribute annotations. In this paper we propose to go beyond pre-recorded benchmarks with post-hoc annotations by presenting an approach that utilizes omnidirectional videos to generate realistic, consistently annotated, short-term tracking scenarios with exactly parameterized motion patterns. We have created an evaluation system, constructed a fully annotated dataset of omnidirectional videos and generators for typical motion patterns. We provide an in-depth analysis of major tracking paradigms which is complementary to the standard benchmarks and confirms the expressiveness of our evaluation approach.

Hierarchical feature learning based on convolutional neural networks (CNN) has recently shown significant potential in various computer vision tasks. While allowing high-quality discriminative feature learning, the downside of CNNs is the lack of explicit structure in features, which often leads to overfitting, absence of reconstruction from partial observations and limited generative abilities. Explicit structure is inherent in hierarchical compositional models, however, these lack the ability to optimize a well-defined cost function. We propose a novel analytic model of a basic unit in a layered hierarchical model with both explicit compositional structure and a well-defined discriminative cost function. Our experiments on two datasets show that the proposed compositional model performs on a par with standard CNNs on discriminative tasks, while, due to explicit modeling of the structure in the feature units, affording a straight-forward visualization of parts and faster inference due to separability of the units.

The next generation service robots are expected to co-exist with humans in their homes. Such a mobile robot requires an efficient representation of space, which should be compact and expressive, for effective operation in real-world environments. In this paper we present a novel approach for 2D ground-plan-like laser-range-data-based room categorization that builds on a compositional hierarchical representation of space, and show how an additional abstraction layer, whose parts are formed by merging partial views of the environment followed by graph extraction, can achieve improved categorization performance. A new algorithm is presented that finds a dictionary of exemplar elements from a multi-category set, based on the affinity measure defined among pairs of elements. This algorithm is used for part selection in new layer construction. Room categorization experiments have been performed on a challenging publicly available dataset, which has been extended in this work. State-of-the-art results were obtained by achieving the most balanced performance over all categories.

A service robot that operates in a previously-unseen home environment should be able to recognize the functionality of the rooms it visits, such as a living room, a bathroom, etc. We present a novel part-based model and an approach for room categorization using data obtained from a visual sensor. Images are represented with sets of unordered parts that are obtained by object-agnostic region proposals, and encoded using state-of-the-art image descriptor extractor - a convolutional neural network (CNN). An approach is proposed that learns category-specific discriminative parts for the part-based model. The proposed approach was compared to the state-of-the-art CNN trained specifically for place recognition. Experimental results show that the proposed approach outperforms the holistic CNN by being robust to image degradation, such as occlusions, modifications of image scaling, and aspect changes. In addition, we report non-negligible annotation errors and image duplicates in a popular dataset for place categorization and discuss annotation ambiguities.

Deformable part models exhibit excellent performance in tracking non-rigidly deforming targets, but are usually outperformed by holistic models when the target does not deform or in the presence of uncertain visual data. The reason is that part-based models require estimation of a larger number of parameters compared to holistic models and since the updating process is self-supervised, the errors in parameter estimation are amplified with time, leading to a faster accuracy reduction than in holistic models. On the other hand, the robustness of part-based trackers is generally greater than in holistic trackers. We address the problem of self-supervised estimation of a large number of parameters by introducing controlled graduation in estimation of the free parameters. We propose decomposing the visual model into several sub-models, each describing the target at a different level of detail. The sub-models interact during target localization and, depending on the visual uncertainty, serve for cross-sub-model supervised updating. A new tracker is proposed based on this model which exhibits the qualities of part-based as well as holistic models. The tracker is tested on the highly-challenging VOT2013 and VOT2014 benchmarks, outperforming the state-of-the-art.

This article presents an integrated robot system capable of interactive learning in dialogue with a human. Such a system needs to have several competencies and must be able to process different types of representations. In this article, we describe a collection of mechanisms that enable integration of heterogeneous competencies in a principled way. Central to our design is the creation of beliefs from visual and linguistic information, and the use of these beliefs for planning system behaviour to satisfy internal drives. The system is able to detect gaps in its knowledge and to plan and execute actions that provide information needed to fill these gaps. We propose a hierarchy of mechanisms which are capable of engaging in different kinds of learning interactions, e.g. those initiated by a tutor or by the system itself. We present the theory these mechanisms are build upon and an instantiation of this theory in the form of an integrated robot system. We demonstrate the operation of the system in the case of learning conceptual models of objects and their visual properties.

This paper addresses the problem of single-target tracker performance evaluation. We consider the performance measures, the dataset and the evaluation system to be the most important components of tracker evaluation and propose requirements for each of them. The requirements are the basis of a new evaluation methodology that aims at a simple and easily interpretable tracker comparison. The ranking-based methodology addresses tracker equivalence in terms of statistical significance and practical differences. A fully-annotated dataset with per-frame annotations with several visual attributes is introduced. The diversity of its visual properties is maximized in a novel way by clustering a large number of videos according to their visual attributes. This makes it the most sophistically constructed and annotated dataset to date. A multi-platform evaluation system allowing easy integration of third-party trackers is presented as well. The proposed evaluation methodology was tested on the VOT2014 challenge on the new dataset and 38 trackers, making it the largest benchmark to date. Most of the tested trackers are indeed state-of-the-art since they outperform the standard baselines, resulting in a highly-challenging benchmark. An exhaustive analysis of the dataset from the perspective of tracking difficulty is carried out. To facilitate tracker comparison a new performance visualization technique is proposed.

Obstacle detection plays an important role in unmanned surface vehicles (USVs). The USVs operate in a highly diverse environments in which an obstacle may be a floating piece of wood, a scuba diver, a pier, or a part of a shoreline, which presents a significant challenge to continuous detection from images taken on board. This paper addresses the problem of online detection by constrained, unsupervised segmentation. To this end, a new graphical model is proposed that affords a fast and continuous obstacle image-map estimation from a single video stream captured on board a USV. The model accounts for the semantic structure of marine environment as observed from USV by imposing weak structural constraints. A Markov random field framework is adopted and a highly efficient algorithm for simultaneous optimization of model parameters and segmentation mask estimation is derived. Our approach does not require computationally intensive extraction of texture features and comfortably runs in real time. The algorithm is tested on a new, challenging, dataset for segmentation, and obstacle detection in marine environments, which is the largest annotated dataset of its kind. Results on this dataset show that our model outperforms the related approaches, while requiring a fraction of computational effort.

The problem of visual tracking evaluation is sporting a large variety of performance measures, and largely suffers from lack of consensus about which measures should be used in experiments. This makes the cross-paper tracker comparison difficult. Furthermore, as some measures may be less effective than others, the tracking results may be skewed or biased toward particular tracking aspects. In this paper, we revisit the popular performance measures and tracker performance visualizations and analyze them theoretically and experimentally. We show that several measures are equivalent from the point of information they provide for tracker comparison and, crucially, that some are more brittle than the others. Based on our analysis, we narrow down the set of potential measures to only two complementary ones, describing accuracy and robustness, thus pushing toward homogenization of the tracker evaluation methodology. These two measures can be intuitively interpreted and visualized and have been employed by the recent visual object tracking challenges as the foundation for the evaluation methodology.

The problem of visual tracking evaluation is sporting an abundance of performance measures, which are used by various authors, and largely suffers from lack of consensus about which measures should be preferred. This is hampering the cross-paper tracker comparison and faster advancement of the field. In this paper we provide an overview of the popular measures and performance visualizations and their critical theoretical and experimental analysis. We show that several measures are equivalent from the point of information they provide for tracker comparison and, crucially, that some are more brittle than the others. Based on our analysis we narrow down the set of potential measures to only two complementary ones that can be intuitively interpreted and visualized, thus pushing towards homogenization of the tracker evaluation methodology.

We propose a new method for a supervised online estimation of probabilistic discriminative models for classification tasks. The method estimates the class distributions from a stream of data in the form of Gaussian mixture models (GMMs). The reconstructive updates of the distributions are based on the recently proposed online kernel density estimator (oKDE). We maintain the number of components in the model low by compressing the GMMs from time to time. We propose a new cost function that measures loss of interclass discrimination during compression, thus guiding the compression toward simpler models that still retain discriminative properties. The resulting classifier thus independently updates the GMM of each class, but these GMMs interact during their compression through the proposed cost function. We call the proposed method the online discriminative kernel density estimator (odKDE). We compare the odKDE to oKDE, batch state-of-the-art kernel density estimators (KDEs), and batch/incremental support vector machines (SVM) on the publicly available datasets. The odKDE achieves comparable classification performance to that of best batch KDEs and SVM, while allowing online adaptation from large datasets, and produces models of lower complexity than the oKDE.

This paper addresses the problem of tracking objects which undergo rapid and significant appearance changes. We propose a novel coupled-layer visual model that combines the target's global and local appearance by interlacing two layers. The local layer in this model is a set of local patches that geometrically constrain the changes in the target's appearance. This layer probabilistically adapts to the target's geometric deformation, while its structure is updated by removing and adding the local patches. The addition of these patches is constrained by the global layer that probabilistically models the target's global visual properties, such as color, shape, and apparent local motion. The global visual properties are updated during tracking using the stable patches from the local layer. By this coupled constraint paradigm between the adaptation of the global and the local layer, we achieve a more robust tracking through significant appearance changes. We experimentally compare our tracker to 11 state-of-the-art trackers. The experimental results on challenging sequences confirm that our tracker outperforms the related trackers in many cases by having a smaller failure rate as well as better accuracy. Furthermore, the parameter analysis shows that our tracker is stable over a range of parameter values.

We present a novel dataset for evaluation of object matching and recognition methods in surveillance scenarios. Dataset consists of more than 23,000 images, depicting 15 persons and nine vehicles. A ground truth data - the identity of each person or vehicle - is provided, along with the coordinates of the bounding box in the full camera image. The dataset was acquired from 36 stationary camera views using a variety of surveillance cameras with resolutions ranging from standard VGA to three megapixel. 27 cameras observed the persons and vehicles in an outdoor environment, while the remaining nine observed the same persons indoors. The activity of persons was planned in advance, they drive the cars to the parking lot, exit the cars and walk around the building, through the main entrance, and up the stairs, towards the first floor of the building. The intended use of the dataset is performance evaluation of computer vision methods that aim to (re)identify people and objects from many different viewpoints in different environments and under variable conditions. Due to variety of camera locations, vantage points and resolutions, the dataset provides means to adjust the difficulty of the identification task in a controlled and documented manner. An interface for easy use of dataset within Matlab is provided as well, and the data is complemented by baseline results using a basic color histogram-based descriptor. While the cropped images of persons and vehicles represent the primary data in our dataset, we also provide full-frame images and a set of tracklets for each object as a courtesy to the dataset users.

Mobile robots need an effective spatial model for the successful operation in real-world environment. The model should be compact and simultaneously possess large expressive power. Moreover, it should scale well. In this paper we propose a new hierarchical representation of space, whose compositional structure is learned based on statistically significant observations. We have focused on a two dimensional space, since many robots perceive their surroundings in two dimensions with the use of a laser range finder or a sonar. We also propose the use of a low-level image descriptor for addressing the room classification problem, by which we demonstrate the performance of our representation. Using only the lower layers of the hierarchy, we obtain state-of-the-art classification results on demanding datasets.

This paper addresses the problem of tracking objects which undergo rapid and significant appearance changes. We propose a novel coupled-layer visual model that combines the target's global and local appearance. The local layer in this model is a set of local patches that geometrically constrain the changes in the target's appearance. This layer probabilistically adapts to the target's geometric deformation, while its structure is updated by removing and adding the local patches. The addition of the patches is constrained by the global layer that probabilistically models target's global visual properties such as color, shape and apparent local motion. The global visual properties are updated during tracking using the stable patches from the local layer. By this coupled constraint paradigm between the adaptation of the global and the local layer, we achieve a more robust tracking through significant appearance changes. Indeed, the experimental results on challenging sequences confirm that our tracker outperforms the related state-of-the-art trackers by having smaller failure rate as well as better accuracy.

In this paper we present representations and mechanisms that facilitate continuous learning of visual concepts in dialogue with a tutor and show the implemented robot system. We present how beliefs about the world are created by processing visual and linguistic information and show how they are used for planning system behaviour with the aim at satisfying its internal drive - to extend its knowledge. The system facilitates different kinds of learning initiated by the human tutor or by the system itself. We demonstrate these principles in the case of learning about object colours and basic shapes.

In this paper, we propose a framework of hierarchical feature distribution for object matching in a network of visual sensors. In our approach, we hierarchically distribute the information in such a way that each individual node maintains only a small amount of information about the objects seen by the network. Nevertheless, this amount is sufficient to efficiently route queries through the network without any degradation of the matching performance. A set of requirements that have to be fulfilled by the object-matching method to be used in such a framework is defined. We provide examples of mapping four well-known, object-matching methods to a hierarchical feature-distribution scheme. The proposed approach was tested on a standard COIL-100 image database and in a basic surveillance scenario using our own distributed network simulator. The results show that the amount of data transmitted through the network can be significantly reduced in comparison to naive feature-distribution schemes such as flooding.

We propose a new method for online estimation of probabilistic discriminative models. The method is based on the recently proposed online Kernel Density Estimation (oKDE) framework which produces Gaussian mixture models and allows adaptation using only a single data point at a time. The oKDE builds reconstructive models from the data, and we extend it to take into account the interclass discrimination through a new distance function between the classifiers. We arrive at an online discriminative Kernel Density Estimator (odKDE). We compare the odKDE to oKDE, batch state-of-the-art KDEs and support vector machine (SVM) on a standard database. The odKDE achieves comparable classification performance to that of best batch KDEs and SVM, while allowing online adaptation, and produces models of lower complexity than the oKDE.

Dimensionality reduction is an important issue in the context of distributed vision systems. Processing of dimensionality reduced data requires far less network resources (e.g., storage space, network bandwidth) than processing of original data. In this paper we explore the performance of the random projection method for distributed smart cameras. In our tests, random projection is compared to principal component analysis in terms of recognition efficiency (i.e., object recognition). The results obtained on the COIL-20 image data set show good performance of the random projection in comparison to the principal component analysis, which requires distribution of a subspace and therefore consumes more resources of the network. This indicates that random projection method can elegantly solve the problem of subspace distribution in embedded and distributed vision systems. Moreover, even without explicit orthogonalization or normalization of random projection transformation subspace, the method achieves good object recognition efficiency.

We propose a new dynamic model which can be used within blob trackers to track the target's center of gravity. A strong point of the model is that it is designed to track a variety of motions which are usually encountered in applications such as pedestrian tracking, hand tracking, and sports. We call the dynamic model a two-stage dynamic model due to its particular structure, which is a composition of two models: a liberal model and a conservative model. The liberal model allows larger perturbations in the target's dynamics and is able to account for motions in between the random-walk dynamics and the nearly constant-velocity dynamics. On the other hand, the conservative model assumes smaller perturbations and is used to further constrain the liberal model to the target's current dynamics. We implement the two-stage dynamic model in a two-stage probabilistic tracker based on the particle filter and apply it to two separate examples of blob tracking: 1) tracking entire persons and 2) tracking of a person's hands. Experiments show that, in comparison to the widely used models, the proposed two-stage dynamic model allows tracking with smaller number of particles in the particle filter (e.g., 25 particles), while achieving smaller errors in the state estimation and a smaller failure rate. The results suggest that the improved performance comes from the model's ability to actively adapt to the target's motion during tracking.