Coding Visual Features Extracted From Video Sequences

Baroffio, Luca; Cesana, Matteo; Redondi, Alessandro E. C.; Tagliasacchi, Marco; Tubaro, Stefano

doi:10.1109/tip.2014.2312617

Cited by 50 publications

(56 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This set comprises 41 subjects. After finding 1 fg11: network class, ht: high-throughput, l3: 3 layers, 1: top-1 best performing instance 2 The aligned dataset can be downloaded at http://www.ic.unicamp.br/ chiachia/resources/pubfig83-aligned the necessary network parameters, this set is completely disregarded and the calculated network and its parameters are then used for feature extraction in the evaluation set. Complementarily, the evaluation set comprises the remaining 42 subjects of the PubFig83 dataset.…”

Section: Methodsmentioning

confidence: 99%

“…Recently, the novel Analyze-Then-Compress (ATC) paradigm has been proposed [21,20,2,17]. In this paradigm, low-power nodes acquire visual content and extract higher-level information from it by means of efficient feature extraction algorithms.…”

Section: Introductionmentioning

confidence: 99%

“…Such content is then compressed resorting to ad-hoc coding primitives and transmitted to a central node, that performs visual analysis. Such approach has been proven to achieve good results in terms of rate-energy-accuracy performance for a number of tasks including content-based retrieval [21,20], object tracking [2] and mobile augmented reality [17]. Moving part of the computational burden, that is, detecting faces and computing compact yet representative features, from a centralized node to sensing devices produces several positive effects:…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Rate-energy-accuracy optimization of convolutional architectures for face recognition

Bondi

Baroffio

Cesana

et al. 2016

Journal of Visual Communication and Image Representation

View full text Add to dashboard Cite

Face recognition systems based on Convolutional Neural Networks (CNNs) or Convolutional architectures currently represent the state of the art, achieving an accuracy comparable to that of humans. Nonetheless, there are two issues that might hinder their adoption on distributed battery-operated devices (e.g., visual sensor nodes, smartphones, and wearable devices). First, Convolutional architectures are usually computationally demanding, especially when the depth of the network is increased to maximize accuracy. Second, transmitting the output features produced by a CNN might require a bitrate higher than the one needed for coding the input image. Therefore, in this paper we address the problem of optimizing the energy-rate-accuracy characteristics of a convolutional architecture for face recognition. We carefully profile a CNN implementation on a Raspberry Pi device and optimize the structure of the neural network, achieving a 17-fold speedup without significantly affecting recognition accuracy. Moreover, we propose a coding architecture custom-tailored to features extracted by such model.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Rate-energy-accuracy optimization of convolutional architectures for face recognition

Bondi

Baroffio

Cesana

et al. 2016

Journal of Visual Communication and Image Representation

View full text Add to dashboard Cite

show abstract

“…For each keypoint, a local feature is produced, summarizing the patch surrounding the keypoint. Since such features can be efficiently compressed and used for automatic analysis, they constitute an excellent candidate for enabling advanced visual tasks in energy-and bandwidthlimited scenarios [6], [7], [8], [9].…”

Section: Introductionmentioning

confidence: 99%

EZ-VSN: An Open-Source and Flexible Framework for Visual Sensor Networks

Bondi

Baroffio

Cesana

et al. 2016

IEEE Internet Things J.

Self Cite

View full text Add to dashboard Cite

Abstract-We present a complete, open-source framework for rapid experimentation of Visual Sensor Networks (VSN) solutions. From the software point of view, we base our architecture on open-source and widely known C++ libraries to provide the basic image processing and networking primitives. The resulting system can be leveraged to create different types of VSNs, characterized by the presence of multiple cameras, relays and cooperator nodes, and can be run on any Linux-based hardware platform, such as the BeagleBone Black. To demonstrate the flexibility of the proposed framework, we describe two different application scenarios typical of VSNs, namely object recognition and parking monitoring. The framework is then used to evaluate the benefits of two complementary paradigms for networked visual analysis recently discussed in the literature. In the traditional Compress-then-Analyze (CTA) paradigm, compressed images are transmitted from camera nodes to a central controller, where they are analyzed. In the novel Analyze-then-Compress (ATC) paradigm, camera nodes extracts and compress local features from the acquired images. Such features are transmitted to the central controller and used to perform visual analysis. We show that the ATC paradigm outperforms CTA from the consumed energy point of view, at the same target analysis accuracy in both the application scenarios.

show abstract

“…Temporally stable detectors and consistent descriptors lead to significant improvements in both accuracy and coding efficiency, especially considering tracking scenarios [9]. As to the compression of features extracted from video, several different lossy and lossless architectures have been proposed, targeting either local [9,10] or global [11] features. Such architectures usually take inspiration from the traditional video coding techniques, adapting the coding process to the signal at hand.…”

mentioning

confidence: 99%