�ݺ�ߣshows by User: galuscakova

�ݺ�ߣshows by User: galuscakova / http://www.slideshare.net/images/logo.gif �ݺ�ߣshows by User: galuscakova / Sun, 12 Jul 2020 13:53:24 GMT �ݺ�ߣShare feed for �ݺ�ߣshows by User: galuscakova Combining Evidence for Cross-language Information Retrieval /slideshow/combining-evidence-for-crosslanguage-information-retrieval/236830928 combiningevidence-200712135324
System combination has been extensively studied in monolingual information retrieval, but the problem is understudied in cross-language retrieval in which queries are expressed in one language, but documents are written in another. One notable characteristic of cross-language retrieval, however, is the potential for a greater diversity of system design, since translation and retrieval components both exhibit substantial design spaces. Due to the large diversity of the systems in cross-language retrieval, the potential range of combinations is orders of magnitude larger than in monolingual applications. I show that evidence combination works well in cross-language retrieval, achieving improvements of 40% relative to the best single system. The best results are obtained using post-retrieval evidence combination, which is able to incorporate many diverse high-quality systems. Because hundreds of different systems can be built, the effectiveness of alternative approaches for managing the complexity is also explored. Both system clustering and expert judgment regarding diversity can help to limit the combinatorial growth of time complexity arising when selections among large numbers of systems need to be made.]]>
System combination has been extensively studied in monolingual information retrieval, but the problem is understudied in cross-language retrieval in which queries are expressed in one language, but documents are written in another. One notable characteristic of cross-language retrieval, however, is the potential for a greater diversity of system design, since translation and retrieval components both exhibit substantial design spaces. Due to the large diversity of the systems in cross-language retrieval, the potential range of combinations is orders of magnitude larger than in monolingual applications. I show that evidence combination works well in cross-language retrieval, achieving improvements of 40% relative to the best single system. The best results are obtained using post-retrieval evidence combination, which is able to incorporate many diverse high-quality systems. Because hundreds of different systems can be built, the effectiveness of alternative approaches for managing the complexity is also explored. Both system clustering and expert judgment regarding diversity can help to limit the combinatorial growth of time complexity arising when selections among large numbers of systems need to be made.]]> Sun, 12 Jul 2020 13:53:24 GMT /slideshow/combining-evidence-for-crosslanguage-information-retrieval/236830928 galuscakova@slideshare.net(galuscakova) Combining Evidence for Cross-language Information Retrieval galuscakova System combination has been extensively studied in monolingual information retrieval, but the problem is understudied in cross-language retrieval in which queries are expressed in one language, but documents are written in another. One notable characteristic of cross-language retrieval, however, is the potential for a greater diversity of system design, since translation and retrieval components both exhibit substantial design spaces. Due to the large diversity of the systems in cross-language retrieval, the potential range of combinations is orders of magnitude larger than in monolingual applications. I show that evidence combination works well in cross-language retrieval, achieving improvements of 40% relative to the best single system. The best results are obtained using post-retrieval evidence combination, which is able to incorporate many diverse high-quality systems. Because hundreds of different systems can be built, the effectiveness of alternative approaches for managing the complexity is also explored. Both system clustering and expert judgment regarding diversity can help to limit the combinatorial growth of time complexity arising when selections among large numbers of systems need to be made. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/combiningevidence-200712135324-thumbnail.jpg?width=120&height=120&fit=bounds" /> System combination has been extensively studied in monolingual information retrieval, but the problem is understudied in cross-language retrieval in which queries are expressed in one language, but documents are written in another. One notable characteristic of cross-language retrieval, however, is the potential for a greater diversity of system design, since translation and retrieval components both exhibit substantial design spaces. Due to the large diversity of the systems in cross-language retrieval, the potential range of combinations is orders of magnitude larger than in monolingual applications. I show that evidence combination works well in cross-language retrieval, achieving improvements of 40% relative to the best single system. The best results are obtained using post-retrieval evidence combination, which is able to incorporate many diverse high-quality systems. Because hundreds of different systems can be built, the effectiveness of alternative approaches for managing the complexity is also explored. Both system clustering and expert judgment regarding diversity can help to limit the combinatorial growth of time complexity arising when selections among large numbers of systems need to be made.

Combining Evidence for Cross-language Information Retrieval from Petra Galuscakova

]]> 111 0 https://cdn.slidesharecdn.com/ss_thumbnails/combiningevidence-200712135324-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Multimodal Features for Linking Television Content /slideshow/multimodal-features-for-linking-television-content/75916528 aplikace-galuscakova-170512102918
How can acoustic and visual features be combined with text-based search methods applied on automatic transcripts and subtitles and help to retrieve television content.]]>
How can acoustic and visual features be combined with text-based search methods applied on automatic transcripts and subtitles and help to retrieve television content.]]> Fri, 12 May 2017 10:29:18 GMT /slideshow/multimodal-features-for-linking-television-content/75916528 galuscakova@slideshare.net(galuscakova) Multimodal Features for Linking Television Content galuscakova How can acoustic and visual features be combined with text-based search methods applied on automatic transcripts and subtitles and help to retrieve television content. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/aplikace-galuscakova-170512102918-thumbnail.jpg?width=120&height=120&fit=bounds" /> How can acoustic and visual features be combined with text-based search methods applied on automatic transcripts and subtitles and help to retrieve television content.

Multimodal Features for Linking Television Content from Petra Galuscakova

]]> 1319 11 https://cdn.slidesharecdn.com/ss_thumbnails/aplikace-galuscakova-170512102918-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Czech Malach Cross-lingual Speech Retrieval Test Collection /slideshow/czech-malach-crosslingual-speech-retrieval-test-collection/59378704 archivesunleashedgaluscakova-160310153918
Presentation of Czech Malach collection of Holocaust survivors testimonies with topical annotations presented at the Archives Unleashed Hackathon.]]>
Presentation of Czech Malach collection of Holocaust survivors testimonies with topical annotations presented at the Archives Unleashed Hackathon.]]> Thu, 10 Mar 2016 15:39:18 GMT /slideshow/czech-malach-crosslingual-speech-retrieval-test-collection/59378704 galuscakova@slideshare.net(galuscakova) Czech Malach Cross-lingual Speech Retrieval Test Collection galuscakova Presentation of Czech Malach collection of Holocaust survivors testimonies with topical annotations presented at the Archives Unleashed Hackathon. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/archivesunleashedgaluscakova-160310153918-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation of Czech Malach collection of Holocaust survivors testimonies with topical annotations presented at the Archives Unleashed Hackathon.

Czech Malach Cross-lingual Speech Retrieval Test Collection from Petra Galuscakova

]]> 851 12 https://cdn.slidesharecdn.com/ss_thumbnails/archivesunleashedgaluscakova-160310153918-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Audio Information for Hyperlinking of TV Content /slideshow/audio-information-for-hyperlinking-of-tv-content/54632788 slam-galuscakova-2015-151102074416-lva1-app6892
We explore the use of audio information in the retrieval of multimedia content.]]>
We explore the use of audio information in the retrieval of multimedia content.]]> Mon, 02 Nov 2015 07:44:16 GMT /slideshow/audio-information-for-hyperlinking-of-tv-content/54632788 galuscakova@slideshare.net(galuscakova) Audio Information for Hyperlinking of TV Content galuscakova We explore the use of audio information in the retrieval of multimedia content. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/slam-galuscakova-2015-151102074416-lva1-app6892-thumbnail.jpg?width=120&height=120&fit=bounds" /> We explore the use of audio information in the retrieval of multimedia content.

Audio Information for Hyperlinking of TV Content from Petra Galuscakova

]]> 637 7 https://cdn.slidesharecdn.com/ss_thumbnails/slam-galuscakova-2015-151102074416-lva1-app6892-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Multimodal Features for Search and Hyperlinking of Video Content /slideshow/idiap-galuscakova/41115261 idiap-galuscakova-141104114409-conversion-gate01
In the talk, I will discuss content-based retrieval in audio-visual collections. I will focus on retrieval of relevant segments of video using a textual query. In addition, I will describe techniques for detecting hyperlinks within audio-visual collections. Our retrieval system ranked first in the MediaEval 2014 Search and Hyperlinking shared task. The experiments were performed on almost 4000 hours of BBC broadcast video. As the segmentation of the recordings shows to be crucial for high-quality video retrieval and hyperlinking, I will focus on segmentation strategies. I will show the possibility of employment of the prosodic and visual information into the segmentation process. Our decision tree-based segmentation proved to outperform fixed-length segmentation which regularly achieves the best results in the retrieval process. Visual and prosodic similarity are also explored in addition to the hyperlinking based on the subtitles and automatic transcripts. The employment of the visual similarity achieves a constant improvement, while the employment of the prosodic similarity shows a small but promising improvement too.]]>
In the talk, I will discuss content-based retrieval in audio-visual collections. I will focus on retrieval of relevant segments of video using a textual query. In addition, I will describe techniques for detecting hyperlinks within audio-visual collections. Our retrieval system ranked first in the MediaEval 2014 Search and Hyperlinking shared task. The experiments were performed on almost 4000 hours of BBC broadcast video. As the segmentation of the recordings shows to be crucial for high-quality video retrieval and hyperlinking, I will focus on segmentation strategies. I will show the possibility of employment of the prosodic and visual information into the segmentation process. Our decision tree-based segmentation proved to outperform fixed-length segmentation which regularly achieves the best results in the retrieval process. Visual and prosodic similarity are also explored in addition to the hyperlinking based on the subtitles and automatic transcripts. The employment of the visual similarity achieves a constant improvement, while the employment of the prosodic similarity shows a small but promising improvement too.]]> Tue, 04 Nov 2014 11:44:09 GMT /slideshow/idiap-galuscakova/41115261 galuscakova@slideshare.net(galuscakova) Multimodal Features for Search and Hyperlinking of Video Content galuscakova In the talk, I will discuss content-based retrieval in audio-visual collections. I will focus on retrieval of relevant segments of video using a textual query. In addition, I will describe techniques for detecting hyperlinks within audio-visual collections. Our retrieval system ranked first in the MediaEval 2014 Search and Hyperlinking shared task. The experiments were performed on almost 4000 hours of BBC broadcast video. As the segmentation of the recordings shows to be crucial for high-quality video retrieval and hyperlinking, I will focus on segmentation strategies. I will show the possibility of employment of the prosodic and visual information into the segmentation process. Our decision tree-based segmentation proved to outperform fixed-length segmentation which regularly achieves the best results in the retrieval process. Visual and prosodic similarity are also explored in addition to the hyperlinking based on the subtitles and automatic transcripts. The employment of the visual similarity achieves a constant improvement, while the employment of the prosodic similarity shows a small but promising improvement too. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/idiap-galuscakova-141104114409-conversion-gate01-thumbnail.jpg?width=120&height=120&fit=bounds" /> In the talk, I will discuss content-based retrieval in audio-visual collections. I will focus on retrieval of relevant segments of video using a textual query. In addition, I will describe techniques for detecting hyperlinks within audio-visual collections. Our retrieval system ranked first in the MediaEval 2014 Search and Hyperlinking shared task. The experiments were performed on almost 4000 hours of BBC broadcast video. As the segmentation of the recordings shows to be crucial for high-quality video retrieval and hyperlinking, I will focus on segmentation strategies. I will show the possibility of employment of the prosodic and visual information into the segmentation process. Our decision tree-based segmentation proved to outperform fixed-length segmentation which regularly achieves the best results in the retrieval process. Visual and prosodic similarity are also explored in addition to the hyperlinking based on the subtitles and automatic transcripts. The employment of the visual similarity achieves a constant improvement, while the employment of the prosodic similarity shows a small but promising improvement too.

Multimodal Features for Search and Hyperlinking of Video Content from Petra Galuscakova

]]> 1515 5 https://cdn.slidesharecdn.com/ss_thumbnails/idiap-galuscakova-141104114409-conversion-gate01-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Evaluácia tematického vyhľadávania v audiovizuálnych nahrávkach /slideshow/prezentacia-cvhgaluscakova-33240954/33240954 prezentacia-cvh-galuscakova-140407153700-phpapp01
4. výročie Centra vizuálnej histórie Malach]]>
4. výročie Centra vizuálnej histórie Malach]]> Mon, 07 Apr 2014 15:37:00 GMT /slideshow/prezentacia-cvhgaluscakova-33240954/33240954 galuscakova@slideshare.net(galuscakova) Evaluácia tematického vyhľadávania v audiovizuálnych nahrávkach galuscakova 4. výročie Centra vizuálnej histórie Malach <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/prezentacia-cvh-galuscakova-140407153700-phpapp01-thumbnail.jpg?width=120&height=120&fit=bounds" /> 4. výročie Centra vizuálnej histórie Malach

Evalu叩cia tematick辿ho vyh直ad叩vania v audiovizu叩lnych nahr叩vkach from Petra Galuscakova

]]> 793 3 https://cdn.slidesharecdn.com/ss_thumbnails/prezentacia-cvh-galuscakova-140407153700-phpapp01-thumbnail.jpg?width=120&height=120&fit=bounds presentation White http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 CUNI at MediaEval 2013 Similar Segments in Social Speech Task /slideshow/ssss-galuscakova2013/33240140 ssss-galuscakova-2013-140407151756-phpapp01
]]>
]]> Mon, 07 Apr 2014 15:17:56 GMT /slideshow/ssss-galuscakova2013/33240140 galuscakova@slideshare.net(galuscakova) CUNI at MediaEval 2013 Similar Segments in Social Speech Task galuscakova <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/ssss-galuscakova-2013-140407151756-phpapp01-thumbnail.jpg?width=120&height=120&fit=bounds" />

CUNI at MediaEval 2013 Similar Segments in Social Speech Task from Petra Galuscakova

]]> 961 5 https://cdn.slidesharecdn.com/ss_thumbnails/ssss-galuscakova-2013-140407151756-phpapp01-thumbnail.jpg?width=120&height=120&fit=bounds presentation White http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Experiments with Segmentation Strategies for Passage Retrieval in Audio-Visual Documents /slideshow/icmr-galuscakova2014/33197465 icmr-galuscakova-2014-140406173830-phpapp01
ICMR 2014]]>
ICMR 2014]]> Sun, 06 Apr 2014 17:38:30 GMT /slideshow/icmr-galuscakova2014/33197465 galuscakova@slideshare.net(galuscakova) Experiments with Segmentation Strategies for Passage Retrieval in Audio-Visual Documents galuscakova ICMR 2014 <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/icmr-galuscakova-2014-140406173830-phpapp01-thumbnail.jpg?width=120&height=120&fit=bounds" /> ICMR 2014

Experiments with Segmentation Strategies for Passage Retrieval in Audio-Visual Documents from Petra Galuscakova

]]> 1496 4 https://cdn.slidesharecdn.com/ss_thumbnails/icmr-galuscakova-2014-140406173830-phpapp01-thumbnail.jpg?width=120&height=120&fit=bounds presentation White http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Česko-slovenský paralelný korpus určený pre preklad medzi blízkymi jazykmi /slideshow/slovko-prezentacia/25236320 slovko-prezentacia-130814060144-phpapp02
Slovko 2011]]>
Slovko 2011]]> Wed, 14 Aug 2013 06:01:44 GMT /slideshow/slovko-prezentacia/25236320 galuscakova@slideshare.net(galuscakova) Česko-slovenský paralelný korpus určený pre preklad medzi blízkymi jazykmi galuscakova Slovko 2011 <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/slovko-prezentacia-130814060144-phpapp02-thumbnail.jpg?width=120&height=120&fit=bounds" /> Slovko 2011

ト憩sko-slovenskテス paralelnテス korpus urト稿nテス pre preklad medzi blテュzkymi jazykmi from Petra Galuscakova

]]> 806 3 https://cdn.slidesharecdn.com/ss_thumbnails/slovko-prezentacia-130814060144-phpapp02-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Application of Topic Segmentation in Audiovisual Information Retrieval /slideshow/i3-galuscakova/25236216 i3galuscakova-130814055714-phpapp02
Week of Doctoral Students 2012]]>
Week of Doctoral Students 2012]]> Wed, 14 Aug 2013 05:57:14 GMT /slideshow/i3-galuscakova/25236216 galuscakova@slideshare.net(galuscakova) Application of Topic Segmentation in Audiovisual Information Retrieval galuscakova Week of Doctoral Students 2012 <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/i3galuscakova-130814055714-phpapp02-thumbnail.jpg?width=120&height=120&fit=bounds" /> Week of Doctoral Students 2012

Application of Topic Segmentation in Audiovisual Information Retrieval from Petra Galuscakova

]]> 1978 5 https://cdn.slidesharecdn.com/ss_thumbnails/i3galuscakova-130814055714-phpapp02-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Penalty Functions for Evaluation Measures of Unsegmented Speech Retrieval /slideshow/galuscakova-clef2012/25236033 galuscakova-clef2012-130814054959-phpapp02
Clef 2012]]>
Clef 2012]]> Wed, 14 Aug 2013 05:49:59 GMT /slideshow/galuscakova-clef2012/25236033 galuscakova@slideshare.net(galuscakova) Penalty Functions for Evaluation Measures of Unsegmented Speech Retrieval galuscakova Clef 2012 <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/galuscakova-clef2012-130814054959-phpapp02-thumbnail.jpg?width=120&height=120&fit=bounds" /> Clef 2012

Penalty Functions for Evaluation Measures of Unsegmented Speech Retrieval from Petra Galuscakova

]]> 710 4 https://cdn.slidesharecdn.com/ss_thumbnails/galuscakova-clef2012-130814054959-phpapp02-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 https://public.slidesharecdn.com/v2/images/profile-picture.png https://cdn.slidesharecdn.com/ss_thumbnails/combiningevidence-200712135324-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/combining-evidence-for-crosslanguage-information-retrieval/236830928 Combining Evidence for... https://cdn.slidesharecdn.com/ss_thumbnails/aplikace-galuscakova-170512102918-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/multimodal-features-for-linking-television-content/75916528 Multimodal Features fo... https://cdn.slidesharecdn.com/ss_thumbnails/archivesunleashedgaluscakova-160310153918-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/czech-malach-crosslingual-speech-retrieval-test-collection/59378704 Czech Malach Cross-lin...