Modern surveillance systems have recently started to employ computer vision algorithms for advanced analysis of the captured video content. Public transportation is one of the domains that may highly benefit from the advances in video analysis. This paper presents a video-based surveillance system that uses a deep neural network based face verification algorithm to accurately and robustly re-identify a subject person. Our implementation is highly scalable due to its container-based architecture and is easily deployable on a cloud platform to support larger processing loads. During the demo, the users will be able to interactively select a target person from pre-recorded surveillance videos and inspect the results on our web-based visualization platform.

With the increasing success of deep learning in various applications, there is an increasing need to have deep models that can be used for deployment in real-time and/or resource constrained scenarios. In this context, this paper analyzes the pruning of deep models for object detection in order to reduce the number of weights and hence the number of computations. Very deep networks based on ResNet like architectures, like YOLOv3 have unique challenges when attempting to prune them. This paper proposes a network pruning technique based on agglomerative clustering for the feature extractor and using mutual information for the detector. The performance of the proposed techniques is also compared with that of a relatively shallow network, i.e., YOLOv2. A compression percentage of around 30% results in a 10% drop of mean average precision (mAP) in YOLOv3, whereas in YOLOv2 the drop was around 6% on the COCO dataset.

Applications like autonomous driving, surveillance, or any application that demands scene analysis requires object detection, semantic segmentation and instance segmentation. In this paper, we focus on the problem of detecting each instance of a specific category of objects, specifically persons. A novel method for object detection is proposed based on a deep counting model. The feature extractor of the deep counting model is extended with additional layers for segmenting specific instances. While the feature extractor of the deep counting model already focuses on the persons in the scene, the segmentation layers help to get a more accurate estimation of the foreground with persons and the instance segmentation is able to estimate separate instances of persons. Our proposed method outperforms other methods on the CUHK08 dataset with an Average Miss Rate (AMR) of 14% and on the PETS09 dataset with an AMR of 41%.

Deep convolutional neural networks (CNNs) have achieved tremendous success in image recognition tasks. However, the performance of CNNs degrade in situations where the input image is degraded by compression artifacts, blur or noise. In this paper, we analyze some of the common CNNs for degradations in images caused by Gaussian noise, blur as well as compression using JPEG and JPEG 2000 for the full range of quality factors. Moreover, we propose a method to improve the performance of CNNs for image classification in the presence of input images with degradations based on a master-slave architecture. Our method was found to perform well for individual and combined degradations.

Pedestrian detection is an important task in surveillance applications and becomes particularly challenging when pedestrians are close together or occluding one another. This paper presents a novel approach to detect pedestrians in such challenging scenarios. A deep convolutional neural network trained for counting is specialized to count one pedestrian. The feature extractor learned thereby is exploited to detect one pedestrian at a time iteratively. For the base counting model and the specialization, extensive annotation efforts are not required since only a single number at the image level is used. Use of our method on pedestrian datasets with occlusion showed an improvement in the average miss rate values as compared to other methods for handling occlusion.

Pedestrian detection is an important task for applications like surveillance, driver assistance systems and autonomous driving. We present a novel approach for detecting pedestrians using a deep convolutional neural network (CNN) trained for counting pedestrians. Our method avoids the need for annotation of the position of the pedestrians in the training data via bounding boxes. The deconvolved outputs of the filters of the trained counting model are used to detect the pedestrians. The average miss rate values on the tested datasets were found to be in the same range as other methods in spite of a simpler training using only pedestrian counts. This method is found to be suitable for detecting pedestrians in crowded scenes with occlusion as well as less crowded scenes.

Counting pedestrians in surveillance applications is a com m n scenario. However, it is often challenging to obtain sufficient annotated training data, especially so fo r creating models using deep learning which require a large amount of training data. To address this problem, thi s paper explores the possibility of training a deep convolutional neural network (CNN) entirely from syntheti cally generated images for the purpose of counting pedestrians. Nuances of transfer learning are exploited to train models from a base model trained for image classification. A direct approach and a hierarchical approa ch are used during training to enhance the capability of the model for counting higher number of pedestrians. The t rained models are then tested on natural images of completely different scenes captured by different acqui sition systems not experienced by the model during training. Furthermore, the effectiveness of the cross entr opy cost function and the squared error cost function are evaluated and analyzed for the scenario where a model is t rained entirely using synthetic images. The performance of the trained model for the test images from the target site can be improved by fine-tuning using the image of the background of the target site.

Pixelwise linear prediction using backward-adaptive least-squares or weighted least-squares estimation of prediction coefficients is currently among the state-of-the-art methods for lossless image compression. While current research is focused on mean intensity prediction of the pixel to be transmitted, best compression requires occurrence probability estimates for all possible intensity values. Apart from common heuristic approaches, we show how prediction error variance estimates can be derived from the (weighted) least-squares training region and how a complete probability distribution can be built based on an autoregressive image model. The analysis of image stationarity properties further allows deriving a novel formula for weight computation in weighted least-squares proofing and generalizing ad hoc equations from the literature. For sparse intensity distributions in non-natural images, a modified image model is presented. Evaluations were done in the newly developed C++ framework volumetric, artificial, and natural image lossless coder (Vanilc), which can compress a wide range of images, including 16-bit medical 3D volumes or multichannel data. A comparison with several of the best available lossless image codecs proofs that the method can achieve very competitive compression ratios. In terms of reproducible research, the source code of Vanilc has been made public.

This paper presents a moving object detection algorithm for H.264/AVC video streams that is applied in the compressed domain. The method is able to extract and analyze several syntax elements from any H.264/AVC-compliant bit stream. The number of analyzed syntax elements depends on the mode in which the method operates. The algorithm is able to perform either a spatiotemporal analysis in a single step or a two-step analysis that starts with a spatial analysis of each frame, followed by a temporal analysis of several subsequent frames. Thereby, in each mode either only (sub-)macroblock types and partition modes or, additionally, quantization parameters are analyzed. The evaluation of these syntax elements enables the algorithm to determine a “weight” for each 4×4 block of pixels that indicates the level of motion within this block. A final segmentation after creating these weights segments each frame to foreground and background and hence indicates the positions and sizes of all moving objects. Our experiments show that the algorithm is able to efficiently detect moving objects in the compressed domain and that it is configurable to process a large number of parallel bit streams in real time.

This paper proposes an approach for the evaluation of DASH-based video transmission from a server located in a Content Delivery Network (CDN) to multiple LTE users. The approach is based on utilizing an analytical model for the HTTP/TCP transmission in the wired/core link, and simulation traces for the packet transmission in the wireless/access link. The core and access link are separately considered by using the Wireless Transmission Control Protocol (WTCP) at the eNodeB, and three scheduling approaches are used for delivering the video packets in the access link. Therefore, in our proposed approach, we emulate DASH-based end-to-end video delivery and observe how the scheduling strategy affects the video streaming quality depending on the number of users in the system.

In this paper, we present a novel moving object detection algorithm for H.264/AVC-compressed video streams. The algorithm does not require full decoding up to the pixel domain but only parsing the compressed bit streams. Thereby, only syntax elements for reconstructing (sub-)macroblock types and quantization parameters are extracted. These features are used to segment the video frames into foreground and background and, according to this segmentation, to identify regions containing moving objects. In a first step, (sub-)macroblock types are analyzed to create initial maps indicating for each block the “weight” for the presence of a moving object. These maps serve as input for our novel spatio-temporal detection algorithm to refine the weight indicating the level of motion for each block. Finally, quantization parameters of macroblocks are used to apply individual thresholds to the block weights to segment the video frames. Experimental results show that our approach efficiently identifies regions containing moving objects and that the presented algorithm is suitable for processing a large number of video streams in parallel.

In this paper we present a new hybrid framework for detecting and tracking persons in surveillance video streams compressed according to the H.264/AVC video coding standard. The framework consists of three stages and operates in both the compressed and the pixel domain of the video. The combination of compressed and pixel domain represents the hybrid character. Its main objective is to significantly reduce the amount of computation required, in particular for frames and image regions with few people present. In its first stage the proposed framework evaluates the header information for each compressed frame in the video sequence, namely the macroblock type information. This results in a coarse binary mask segmenting the frame into foreground and background. Only the foreground regions are processed further in the second stage that searches for persons in the image pixel domain by applying a person detector based on the Implicit Shape Model. The third stage segments each detected person further with a newly developed method that fuses information from the first two stages. This helps obtaining a finer segmentation for calculating a color histogram suitable for tracking the person using the mean shift algorithm. The proposed framework was experimentally evaluated on a publicly available test set. The results demonstrate that the proposed framework reliably separates frames with and without persons such that the computational load is significantly reduced while the detection performance is kept.

We present a novel low-complexity algorithm for feature-based block size decision in HEVC intra coding. Our approach evaluates a set of point pairs within a coding unit (CU) in order to determine whether a CU should be further split into smaller sub CUs for coding. We apply a modified version of the Binary Robust Independent Elementary Feature (BRIEF) descriptor, which in its original version is usually applied to describe local image properties in the context of image/video analysis. While the elements of the original BRIEF descriptor describe which pixel of a pixel pair has the higher value, the modified descriptor evaluates whether the difference between the pixel values is above a pre-defined threshold, which is determined by training. If a certain number of pixel pair differences exceed their corresponding threshold the CU is split into its four sub CUs. Furthermore, we restrict the feature point pairs to be located within different potential sub CUs. For an adaptive training approach, we achieved in our experiments an average encoding time reduction of 58% compared to the HEVC reference software HM12.0 with an average rate increase of 3.8% at equal quality and an average encoding time reduction of 65% with an average rate increase of 5.98% at equal quality for an offline training approach.

The management of medical emergency, in particular cardiac emergency, requests prompt intervention and the possibility to communicate in real time from the emergency area / ambulance to the hospital as much diagnostic information as possible about the patient. This would enable a prompt emergency diagnosis and operation and the possibility to prepare the appropriate actions in the suitable hospital department. To address this scenario, the CONCERTO European project proposed a wireless communication system based on a novel cross-layer architecture, including the integration of building blocks for medical media content fusion, delivery and access. This paper describes the proposed system architecture, outlining the developed components and mechanisms, and the evaluation of the proposed system, carried out in a hospital with the support of medical staff. The technical results and the feedback received highlight the impact of the CONCERTO approach in the healthcare domain, in particular in enabling a prompt and reliable diagnosis in challenging medical emergency scenarios.