Efficient processing of high-resolution video streams is safety-critical for many robotics applications such as autonomous driving. Image downsampling is a commonly adopted technique to ensure the latency constraint is met. However, this naive approach greatly restricts an object detector’s capability to identify small objects. In this paper, we propose an attentional approach that elastically magnifies certain regions while maintaining a small input canvas. The magnified regions are those that are believed to have a high probability of containing an object, whose signal can come from a dataset-wide prior or frame-level prior computed from recent object predictions. The magnification is implemented by a KDE-based mapping to transform the bounding boxes into warping parameters, which are then fed into an image sampler with anti-cropping regularization. The detector is then fed with the warped image and we apply a differentiable backward mapping to get bounding box outputs in the original space. Our regional magnification allows algorithms to make better use of high-resolution input without incurring the cost of high-resolution processing. On the autonomous driving datasets Argoverse-HD and BDD100K, we show our proposed method boosts the detection AP over standard Faster R-CNN, with and without finetuning. Additionally, building on top of the previous state-of-the-art in streaming detection, our method sets a new record for streaming AP on Argoverse-HD (from 17.8 to 23.0 on a GTX 1080 Ti GPU), suggesting that it has achieved a superior accuracy-latency tradeoff.

In this paper, we address the problem of automatically assembling shredded documents. We propose a two-step algorithmic framework. First, we digitize each fragment of a given document and extract shape- and content-based local features. Based on these multimodal features, we identify pairs of corresponding points on all pairs of fragments using an SVM classifier. Each pair is considered a point of attachment for aligning the respective fragments. In order to restore the layout of the document, we create a document graph in which nodes represent fragments and edges correspond to alignments. We assign weights to the edges by evaluating the alignments using a set of inter-fragment constraints which take into account shape- and content-based information. Finally, we use an iterative algorithm that chooses the edge having the highest weight during each iteration. However, since selecting edges corresponds to combining groups of fragments and thus provides new evidence, we reevaluate the edge weights after each iteration. We quantitatively evaluate the effectiveness of our approach by conducting experiments on a novel dataset. It comprises a total of 120 pages taken from two magazines which have been shredded and annotated manually. We thus provide the means for a quantitative evaluation of assembly algorithms which, to the best of our knowledge, has not been done before.

This work addresses two challenges in combination: learning with a very limited number of labeled training examples (active learning) and learning in the presence of multiple views for each object where the global model to be learned is spread out over some or all of these views (learning in parallel universes). We propose a new active learning approach which selects the best samples to query the label with the goal of improving overall model accuracy and determining which universe contributes most to the local model. The resulting combination and class-specific weighting of universes provides a significantly better classification accuracy than traditional active learning methods.

When we think of an object in a supervised learning setting, we usually perceive it as a collection of fixed attribute values. Although this setting may be suited well for many classification tasks, we propose a new object representation and therewith a new challenge in data mining: an object is no longer described by one set of attributes but is represented in a hierarchy of attribute sets in different levels of quality. Obtaining a more detailed representation of an object comes with a cost. This raises the interesting question of which objects we want to enhance under a given budget and cost model. This new setting is very useful whenever resources like computing power, memory or time are limited. We propose a new Active Adaptive Algorithm (AAA) to improve objects in an iterative fashion. We demonstrate how to create a hierarchical object representation and prove the effectiveness of our new selection algorithm on these datasets.