Sabine Sternig

Classifier grids for robust adaptive object detection

In this paper we present an adaptive but robust object detector for static cameras by introducing classifier grids. Instead of using a sliding window for object detection we propose to train a separate classifier for each image location, obtaining a very specific object detector with a low false alarm rate. For each classifier corresponding to a grid element we estimate two generative representations in parallel, one describing the objects class and one describing the background. These are combined in order to obtain a discriminative model. To enable to adapt to changing environments these classifiers are learned on-line (i.e., boosting). Continuously learning (24 hours a day, 7 days a week) requires a stable system. In our method this is ensured by a fixed object representation while updating only the representation of the background. We demonstrate the stability in a long-term experiment by running the system for a whole week, which shows a stable performance over time. In addition, we compare the proposed approach to state-of-the-art methods in the field of person and car detection. In both cases we obtain competitive results.

Hough regions for joining instance localization and segmentation

Object detection and segmentation are two challenging tasks in computer vision, which are usually considered as independent steps. In this paper, we propose a framework which jointly optimizes for both tasks and implicitly provides detection hypotheses and corresponding segmentations. Our novel approach is attachable to any of the available generalized Hough voting methods. We introduce Hough Regions by formulating the problem of Hough space analysis as Bayesian labeling of a random field. This exploits provided classifier responses, object center votes and low-level cues like color consistency, which are combined into a global energy term. We further propose a greedy approach to solve this energy minimization problem providing a pixel-wise assignment to background or to a specific category instance. This way we bypass the parameter sensitive non-maximum suppression that is required in related methods. The experimental evaluation demonstrates that state-of-the-art detection and segmentation results are achieved and that our method is inherently able to handle overlapping instances and an increased range of articulations, aspect ratios and scales.

Multi-camera multi-object tracking by robust hough-based homography projections

Recently, several approaches have been introduced for incorporating the information from multiple cameras to increase the robustness of tracking. This allows to handle problems of mutually occluding objects - a reasonable scenario for many tasks such as visual surveillance or sports analysis. However, these methods often ignore problems such as inaccurate geometric constraints and violated geometric assumptions, requiring complex methods to resolve the resulting errors. In this paper, we introduce a new multiple camera tracking approach that inherently avoids these problems. We build on the ideas of generalized Hough voting and extend it to the multiple camera domain. This offers the following advantages: we reduce the amount of data in voting and are robust to projection errors. Moreover, we show that using additional geometric information can help to train more specific classifiers drastically improving the tracking performance. We confirm these findings by comparing our approach to existing (multi-camera) tracking methods.

Multimodal highway monitoring for robust incident detection

We present detection and tracking methods for highway monitoring based on video and audio sensors, and the combination of these two modalities. We evaluate the performance of the different systems on realistic data sets that have been recorded on Austrian highways. It is shown that we can achieve a very good performance for video-based incident detection of wrong-way drivers, still standing vehicles, and traffic jams. Algorithms for simultaneous vehicle and driving direction detection using microphone arrays were evaluated and also showed a good performance on these tasks. Robust tracking in case of difficult weather conditions is achieved through multimodal sensor fusion of video and audio sensors.

Context-driven clustering by multi-class classification in an active learning framework

Tracking and detection of objects often require to apply complex models to cope with the large intra-class variability of the foreground as well as the background class. In this work, we reduce the complexity of a binary classification problem by a context-driven approach. The main idea is to use a hidden multi-class representation to capture multi-modalities in the data finally providing a binary classifier. We introduce virtual classes generated by a context-driven clustering, which are updated using an active learning strategy. By further using an on-line learner the classifier can easily be adapted to changing environmental conditions. Moreover, by adding additional virtual classes more complex scenarios can be handled. We demonstrate the approach for tracking as well as detection on different scenarios reaching state-of-the-art results.

TransientBoost: On-line boosting with transient data

For on-line learning algorithms, which are applied in many vision tasks such as detection or tracking, robust integration of unlabeled samples is a crucial point. Various strategies such as self-training, semi-supervised learning and multiple-instance learning have been proposed. However, these methods are either too adaptive, which causes drifting, or biased by a prior, which hinders incorporation of new (orthogonal) information. Therefore, we propose a new on-line learning algorithm (TransientBoost), which is highly adaptive but still robust. This is realized by using an internal multi-class representation and modeling reliable and unreliable data in separate classes. Unreliable data is considered transient, hence we use highly adaptive learning parameters to adapt to fast changes in the scene while errors fade out fast. In contrast, the reliable data is preserved completely and not harmed by wrong updates. We demonstrate our algorithm on two different tasks, i.e., object detection and object tracking showing that we can handle typical problems considerable better than existing approaches. To demonstrate the stability and the robustness, we show long-term experiments for both tasks.

Learning of Scene-Specific Object Detectors by Classifier Co-Grids

Recently, classifier grids have shown to be a considerablealternative to sliding window approaches for objectdetection from static cameras. The main drawback of suchmethods is that they are biased by the initial model. In fact,the classifiers can be adapted to changing environmentalconditions but due to conservative updates no new objectspecificinformation is acquired. Thus, the goal of this workis to increase the recall of scene-specific classifiers whilepreserving their accuracy and speed. In particular, we introducea co-training strategy for classifier grids using arobust on-line learner. Thus, the robustness is preservedwhile the recall can be increased. The co-training strategyrobustly provides negative as well as positive updates. Inaddition, the number of negative updates can be drasticallyreduced, which additionally speeds up the system. In theexperimental results these benefits are demonstrated on differentpublicly available surveillance benchmark data sets.

On-line inverse multiple instance boosting for classifier grids

Highlights ► Classifier grids showed excellent detection results for stationary cameras. ► On-line adaptive classifiers reduce the complexity of the classification task. ► Fixed update strategies allow long-term stability. ► Short-term stability through proposed inverse multiple instance learning.

Inverse Multiple Instance Learning for Classifier Grids

Recently, classifier grids have shown to be a considerable alternative for object detection from static cameras. However, one drawback of such approaches is drifting if an object is not moving over a long period of time. Thus, the goal of this work is to increase the recall of such classifiers while preserving their accuracy and speed. In particular, this is realized by adapting ideas from Multiple Instance Learning within a boosting framework. Since the set of positive samples is well defined, we apply this concept to the negative samples extracted from the scene: Inverse Multiple Instance Learning. By introducing temporal bags, we can ensure that each bag contains at least one sample having a negative label, providing the required stability. The experimental results demonstrate that using the proposed approach state-of-the-art detection results can by obtained, however, showing superior classification results in presence of non-moving objects.

Hough Forests Revisited: An Approach to Multiple Instance Tracking from Multiple Cameras

Tracking multiple objects in parallel is a difficult task, especially if instances are interacting and occluding each other. To alleviate the arising problems multiple camera views can be taken into account, which, however, increases the computational effort. Evoking the need for very efficient methods, often rather simple approaches such as background subtraction are applied, which tend to fail for more difficult scenarios. Thus, in this work, we introduce a powerful multi-instance tracking approach building on Hough Forests. By adequately refining the time consuming building blocks, we can drastically reduce their computational complexity without a significant loss in accuracy. In fact, we show that the test time can be reduced by one to two orders of magnitude, allowing to efficiently process the large amount of image data coming from multiple cameras. Furthermore, we adapt the pre-trained generic forest model in an online manner to train an instance-specific model, making it well suited for multi-instance tracking. Our experimental evaluations show the effectiveness of the proposed efficient Hough Forests for object detection as well as for the actual task of multi-camera tracking.