Year: 2013

Video Matching

Amr Ahmed Uncategorized

Video Matching Using DC-image and Local Features Saddam Bekhet, Amr Ahmed, Andrew Hunter

Paper (PDF): Video Matching Using DC-image and Local Features – WCE2013_pp2209-2214

Video Demos:  … soon..

Poster for the WCE'13 paper that was awarded "Best Student Paper" award,
Poster for the WCE’13 paper that was awarded “Best Student Paper” award,

INTRODUCTION:

Videos, especially the compressed ones, became a major part of our daily life. With the amount of videos growing exponentially, Scientists are being pushed to develop robust tools that could efficiently index and retrieve videos in a way similar to human perception of similarity.

*Based on http://www.youtube.com/yt/press/statistics.html

 

PROBLEM

§Manual annotation is a hard work and annotations are not always available for utilization.
§We need more smarter tagging process for videos.
§With the increasing of compressed videos, more efficient techniques are required to work directly on compressed files, without need for decompression.
AIM
Our aim is to build a framework that will operate on compressed videos (utilizing the DC-images sequence),
CONCLUSION

•DC-IMAGE is suitable for cheaper computations and could be used as basic building block for real-time processing.
•Local features proved to be effective on DC-image, after our introduced modification.

Dr Amjad Altadmri – Graduation Ceremony

Amr Ahmed Computer Visions, dcapi, DCAPI blog, Research, Video Analysis , , , , ,

Congratulations to Dr Amjad Altadmri for completeing his PhD degree. Amjad received his PhD degree in the formal September Graduation Ceremony at the Lincoln Cathedral.

Amjad Graduation Ceremony - September 2013
Amjad Graduation Ceremony – September 2013

 

His PhD titled “Semantic Annotation of Domain-Independent Uncontrolled Videos, Incorporating Visual Similarity and Commonsesne Knowledge Bases”. The work produced a Framework for semantic video annotation. In addition, VisualNet was also produced, which is a semantic Network for Visual-related applications.

The photo shows Dr Amjad Altadmri (Left) with his Director of Studies/Supervisor Dr Amr Ahmed ( right).

A list of publications of Amjad’s PhD are below:

 

Amjad has also participated, with Amr and other members of the DCAPI group,  in various workshops especially the V&L EPSRC Network workshops. They presented sessions and showed posters; see related blog posts:

 

Best Student Paper Award 2013 – WCE 2013

Amr Ahmed Computer Visions, Conference, dcapi, DCAPI blog, Research, Video Analysis , , , , , , , ,

Congratulations to Saddam Bekhet (PhD Researcher) who achieved the “Best Student Paper Award 2013″  for his conference paper entitled “Video Matching Using DC-image and Local Features ” presented earlier in “World Congress on Engineering 2013“ in London .
Award copy20130704_112542

 

 

Abstract: This paper presents a suggested framework for video matching based on local features extracted from the DC-image of MPEG compressed videos, without decompression. The relevant arguments and supporting evidences are discussed for developing video similarity techniques that works directly on compressed videos, without decompression, and especially utilising small size images. Two experiments are carried to support the above. The first is comparing between the DC-image and I-frame, in terms of matching performance and the corresponding computation complexity. The second experiment compares between using local features and global features in video matching, especially in the compressed domain and with the small size images. The results confirmed that the use of DC-image, despite its highly reduced size, is promising as it produces at least similar (if not better) matching precision, compared to the full I-frame. Also, using SIFT, as a local feature, outperforms precision of most of the standard global features. On the other hand, its computation complexity is relatively higher, but it is still within the real-time margin. There are also various optimisations that can be done to improve this computation complexity.

 

Conference paper presented WCE’13 – 3rd July 2013 – London

Amr Ahmed Computer Visions, Conference, dcapi, DCAPI blog, International, Research, Video Analysis , , ,

The paper (titled “Video Matching Using DC-image and Local Features”) was presented by Saddam Bekhet (PhD Rsearcher) in the International Conference of Signal and Image Engineering (ICSIE’13), during the World Congress on Engineering 2013, in London UK.

Abstract:

This paper presents a suggested framework for video matching based on local features extracted from the DC-image of MPEG compressed videos, without decompression. The relevant arguments and supporting evidences are discussed for developing video similarity techniques that works directly on compressed videos, without decompression, and especially utilising small size images. Two experiments are carried to support the above. The first is comparing between the DC-image and I-frame, in terms of matching performance and the corresponding computation complexity. The second experiment compares between using local features and global features in video matching, especially in the compressed domain and with the small size images. The results confirmed that the use of DC-image, despite its highly reduced size, is promising as it produces at least similar (if not better) matching precision, compared to the full I-frame. Also, using SIFT, as a local feature, outperforms precision of most of the standard global features. On the other hand, its computation complexity is relatively higher, but it is still within the real-time margin. There are also various optimisations that can be done to improve this computation complexity.

 

Well done Saddam.

Conference paper Accepted to the “World Congress on Engineering”

Amr Ahmed Computer Visions, Conference, dcapi, DCAPI blog, International, Research, Video Analysis , , ,

 New Conference paper accepted for publishing in  “World Congress on Engineering 2013“.

The paper title is “Video Matching Using DC-image and Local Features ”

Abstract:

This paper presents a suggested framework for video matching based on local features extracted from the DC-image of MPEG compressed videos, without decompression. The relevant arguments and supporting evidences are discussed for developing video similarity techniques that works directly on compressed videos, without decompression, and especially utilising small size images. Two experiments are carried to support the above. The first is comparing between the DC-image and I-frame, in terms of matching performance and the corresponding computation complexity. The second experiment compares between using local features and global features in video matching, especially in the compressed domain and with the small size images. The results confirmed that the use of DC-image, despite its highly reduced size, is promising as it produces at least similar (if not better) matching precision, compared to the full I-frame. Also, using SIFT, as a local feature, outperforms precision of most of the standard global features. On the other hand, its computation complexity is relatively higher, but it is still within the real-time margin. There are also various optimisations that can be done to improve this computation complexity.