Rapid estimation of camera motion from compressed video with application to video annotation

Tan, Yap‐Peng; Saur, D. D.; Kulkami, Siddhi; Ramadge, Peter J.

doi:10.1109/76.825867

Cited by 174 publications

(6 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Instead of using content classification as a pre-processing step before streaming the video sequence, content information can also be extracted from the encoded bitstream [65], [66]. In [67], for example, the amount of details and motion in a video sequence is determined based on the encoded frame sizes of I-, P-and B-pictures.…”

Section: B Modeling Packet Loss Visibilitymentioning

confidence: 99%

No-Reference Bitstream-Based Visual Quality Impairment Detection for High Definition H.264/AVC Encoded Video Sequences

Staelens

Wallendael

Crombecq

et al. 2012

IEEE Trans. on Broadcast.

View full text Add to dashboard Cite

Abstract-Ensuring and maintaining adequate Quality of Experience towards end-users are key objectives for video service providers, not only for increasing customer satisfaction but also as service differentiator. However, in the case of High Definition video streaming over IP-based networks, network impairments such as packet loss can severely degrade the perceived visual quality. Several standard organizations have established a minimum set of performance objectives which should be achieved for obtaining satisfactory quality. Therefore, video service providers should continuously monitor the network and the quality of the received video streams in order to detect visual degradations. Objective video quality metrics enable automatic measurement of perceived quality. Unfortunately, the most reliable metrics require access to both the original and the received video streams which makes them inappropriate for real-time monitoring. In this article, we present a novel no-reference bitstream-based visual quality impairment detector which enables real-time detection of visual degradations caused by network impairments. By only incorporating information extracted from the encoded bitstream, network impairments are classified as visible or invisible to the end-user. Our results show that impairment visibility can be classified with a high accuracy which enables real-time validation of the existing performance objectives.

show abstract

Section: B Modeling Packet Loss Visibilitymentioning

confidence: 99%

No-Reference Bitstream-Based Visual Quality Impairment Detection for High Definition H.264/AVC Encoded Video Sequences

Staelens

Wallendael

Crombecq

et al. 2012

IEEE Trans. on Broadcast.

View full text Add to dashboard Cite

show abstract

“…As such, this classification was performed as a pre-processing step and signaled inside the bit stream itself. Using methods described in [13] and [14] it is however possible to derive motion characteristics from the compressed video bit stream.…”

Section: Constructing and Evaluating Decision Trees For Visual Impairmentioning

confidence: 99%

ViQID: A No-Reference bit stream-based visual quality impairment detector

Staelens

Vercammen

Dhondt

et al. 2010

2010 Second International Workshop on Quality of Multimedia Experience (QoMEX)

View full text Add to dashboard Cite

In order to ensure adequate quality towards the end users at all time, video service providers are getting more interested in monitoring their video streams. Objective video quality metrics provide a means of measuring (audio)visual quality in an automated manner. Unfortunately, most of the current existing metrics cannot be used for real-time monitoring due to their dependencies on the original video sequence. In this paper we present a new objective video quality metric which classifies packet loss as visible or invisible based on information extracted solely from the captured encoded H.264/AVC video bit stream. Our results show that the visibility of packet loss can be predicted with a high accuracy, without the need for deep packet inspection. This enables service providers to monitor quality in real-time.

show abstract

“…Other works were studied the event extraction task from sports data based on web casting and broadcast video [Tan et al (2000), Xu et al (2006Xu et al ( , 2004, Zhang et al (2007)].…”

Section: Event Extractionmentioning

confidence: 99%

Event-Based Recognition of Lived Experiences in User Reviews

Hassan¹,

Buscaldi²,

Gangemi³

2016

Lecture Notes in Computer Science

View full text Add to dashboard Cite

The quantity of user-generated content on the Web is constantly growing at a fast pace.A great share of this content is made of opinions and reviews on products and services.This electronic word-of-mouth is also an important factor in decisions about purchasing these products or services. Users tend to trust other users, especially if they can compare themselves to those who wrote the reviews, or, in other words, they are condent to share some characteristics. For instance, families will prefer to travel in places that have been recommended by other families. We assume that reviews that contain lived experiences are more valuable, since experiences give to the reviews a more subjective cut, allowing readers to project themselves into the context of the writer.With this hypothesis in mind, in this thesis we aim to identify, extract, and represent reported lived experiences in customer reviews by hybridizing Knowledge Extraction and Natural Language Processing techniques in order to accelerate the decision process. For this, we dene a lived user experience as an event mentioned in a review, where the author is among the participants. This denition considers that mentioned events in the text are the most important elements in lived experiences: all lived experiences are based on events, which on turn are clearly dened in time and space. Therefore, we propose an approach to extract events from user reviews, which constitute the basis of an event-based system to identify and extract lived experiences.For the event extraction approach, we transform user reviews into their semantic representations using machine reading techniques. We perform a deep semantic parsing of reviews, detecting the linguistic frames that capture complex relations expressed in the reviews. The event-based lived experience system is carried out in three steps. The rst step operates an event-based review ltering, which identies reviews that may contain lived experiences. The second step consists of extracting relevant events together with their participants. The last step focuses on representing extracted lived experiences in each review as an event sub-graph.In order to test our hypothesis, we carried out some experiments to verify whether lived experiences can be considered as triggers for the ratings expressed by users. Therefore, we used lived experiences as features in a classication system, comparing with the ratings of the reviews in a dataset extracted and manually annotated from Tripadvisor. The results show that lived experiences are actually correlated with the ratings.In conclusion, this thesis provides some interesting contributions in the eld of opinion mining. First of all, the successful application of machine reading to identify lived experiences. Second, the conrmation that lived experiences are correlated to ratings. Finally, the dataset produced to test our hypothesis constitutes also an important contribution of the thesis.

show abstract

Rapid estimation of camera motion from compressed video with application to video annotation

Cited by 174 publications

References 16 publications

No-Reference Bitstream-Based Visual Quality Impairment Detection for High Definition H.264/AVC Encoded Video Sequences

No-Reference Bitstream-Based Visual Quality Impairment Detection for High Definition H.264/AVC Encoded Video Sequences

ViQID: A No-Reference bit stream-based visual quality impairment detector

Event-Based Recognition of Lived Experiences in User Reviews

Contact Info

Product

Resources

About