Scene Understanding: A Survey to See the World at a Single Glance

Pawar, Prajakta; Devendran, V.

doi:10.1109/icct46177.2019.8969051

Cited by 13 publications

(8 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, novel task-driven feature decoders are required to extract descriptive features and discriminative features from the coded measurements and then feed them into corresponding modules of SCV tasks. Afterwards, with the extracted descriptive features, we can further retrieve the qualitative semantic information through scene understanding [30] or video captioning [31] methods. In addition, the discriminative features can be used to retrieve quantitative semantic information by object objection/tracking [23] and distance/velocity estimation approaches [67,68].…”

Section: Measurement Domain Scvmentioning

confidence: 99%

“…Equipped with the rapid development of deep learning, we have made significant progress in the field of SCV. For instance, we are constantly pushing forward the intelligent level of our automatic information processing systems from image classification [19], semantic segmentation [27], object detection and tracking [23] to action recognition [28], facial expression recognition [29], scene understanding [30] and video captioning [31].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

From compressive sampling to compressive tasking: retrieving semantics in compressed domain with low bandwidth

et al. 2022

View full text Add to dashboard Cite

High-throughput imaging is highly desirable in intelligent analysis of computer vision tasks. In conventional design, throughput is limited by the separation between physical image capture and digital post processing. Computational imaging increases throughput by mixing analog and digital processing through the image capture pipeline. Yet, recent advances of computational imaging focus on the “compressive sampling”, this precludes the wide applications in practical tasks. This paper presents a systematic analysis of the next step for computational imaging built on snapshot compressive imaging (SCI) and semantic computer vision (SCV) tasks, which have independently emerged over the past decade as basic computational imaging platforms. SCI is a physical layer process that maximizes information capacity per sample while minimizing system size, power and cost. SCV is an abstraction layer process that analyzes image data as objects and features, rather than simple pixel maps. In current practice, SCI and SCV are independent and sequential. This concatenated pipeline results in the following problems: i) a large amount of resources are spent on task-irrelevant computation and transmission, ii) the sampling and design efficiency of SCI is attenuated, and iii) the final performance of SCV is limited by the reconstruction errors of SCI. Bearing these concerns in mind, this paper takes one step further aiming to bridge the gap between SCI and SCV to take full advantage of both approaches. After reviewing the current status of SCI, we propose a novel joint framework by conducting SCV on raw measurements captured by SCI to select the region of interest, and then perform reconstruction on these regions to speed up processing time. We use our recently built SCI prototype to verify the framework. Preliminary results are presented and the prospects for a joint SCI and SCV regime are discussed. By conducting computer vision tasks in the compressed domain, we envision that a new era of snapshot compressive imaging with limited end-to-end bandwidth is coming.

show abstract

Section: Measurement Domain Scvmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

From compressive sampling to compressive tasking: retrieving semantics in compressed domain with low bandwidth

et al. 2022

View full text Add to dashboard Cite

show abstract

“…A number of papers have shown that recognizing a scene involves understanding the visuals (Ali et al, 2017), object detection (Zhou X. et al, 2017) and estimating geometric features. A first scene classification divides them into object-centered and scene-centered (Pawar and Devendran, 2019). An important result in feature extraction reached 70% accuracy on Sun397dataset, it uses Places-CNNs and ImageNet-CNNs for feature extraction (Herranz et al, 2016).…”

Section: Scene Classification Recognition and Understandingmentioning

confidence: 99%

“…Dahua Lin and Jianxiog Xio documented another model (Lin and Xio, 2013) which uses the geometrical pixel arrangement for semantic interpretation and segmentation (Pawar and Devendran, 2019). The model creates structural layers for outdoor scenes with notable experimental results for semantic segmentation and scene classification.…”

Section: Scene Classification Recognition and Understandingmentioning

confidence: 99%

“…This can be a hint to identify some spatial attributes specific to a scene. The method uses very large datasets, like LSUN which contains millions of images, objects, subjects and scenes (Pawar and Devendran, 2019). 25 years back, in 1995, image understanding was defined as a verbal description of the image (Ralescu et al, 1995).…”

Section: Scene Classification Recognition and Understandingmentioning

confidence: 99%

See 1 more Smart Citation

Short Literature Review for Visual Scene Understanding

Achirei

2021

Bulletin of the Polytechnic Institute of Iași. Electrical Engineering, Power Engineering, Electronics Section

View full text Add to dashboard Cite

Individuals are highly accurate for visually understanding natural scenes. By extracting and extrapolating data we reach the highest stage of scene understanding. In the past few years it proved to be an essential part in computer vision applications. It goes further than object detection by bringing machine perceiving closer to the human one: integrates meaningful information and extracts semantic relationships and patterns. Researchers in computer vision focused on scene understanding algorithms, the aim being to obtain semantic knowledge from the environment and determine the properties of objects and the relations between them. For applications in robotics, gaming, assisted living, augmented reality, etc a fundamental task is to be aware of spatial position and capture depth information. First part of this paper focuses on deep learning solutions for scene recognition following the main leads: low-level features and object detection. In the second part we present extensively the most relevant datasets for visual scene understanding. We take into consideration both directions having in mind future applications.

show abstract