Dynamic attention priors: a new and efficient concept for improving object detection
Recent psychophysical evidence in humans suggests that visual attention is a highly dynamic and predictive process involving precise models of object trajectories. We present a proof-of-concept that such predictive spatial attention can benefit a technical system solving a challenging visual object...
Gespeichert in:
| Veröffentlicht in: | Neurocomputing (Amsterdam) 2016-07, Vol.197, p.14-28 |
|---|---|
| Hauptverfasser: | , , , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 28 |
|---|---|
| container_issue | |
| container_start_page | 14 |
| container_title | Neurocomputing (Amsterdam) |
| container_volume | 197 |
| creator | Gepperth, Alexander R.T. Ortiz, Michael Garcia Sattarov, Egor Heisele, Bernd |
| description | Recent psychophysical evidence in humans suggests that visual attention is a highly dynamic and predictive process involving precise models of object trajectories. We present a proof-of-concept that such predictive spatial attention can benefit a technical system solving a challenging visual object detection task. To this end, we introduce a Bayes-like integration of the so-called dynamic attention priors (DAPs) and dense detection likelihoods, which get enhanced at predicted object positions obtained by the extrapolation of trajectories.
Using annotated video sequences of pedestrians in a parking lot setting, we quantitatively show that DAPs can improve detection performance significantly as compared to a baseline condition relying purely on pattern analysis. |
| doi_str_mv | 10.1016/j.neucom.2016.01.036 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_hal_p</sourceid><recordid>TN_cdi_hal_primary_oai_HAL_hal_01418128v1</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><els_id>S0925231216001259</els_id><sourcerecordid>1816089393</sourcerecordid><originalsourceid>FETCH-LOGICAL-c489t-1f08c4c862b937cdaf11808fc6b21fe4fc5d4e38b8291bd68bffee05ed77d67a3</originalsourceid><addsrcrecordid>eNp9kE1PxCAQhonRxPXjH3jgqIdWhnZb6sHE-J1s4kXPhMKgbLawQneN_142NR49TYBn3hkeQs6AlcCguVyWHjc6DCXPp5JByapmj8xAtLwQXDT7ZMY6Pi94BfyQHKW0ZAxa4N2MvN59ezU4TdU4oh9d8HQdXYjpiirq8Ysqbyha67TLz1QHr3E9UhsidcM6hq3z7zT0S9QjNTjmkiNOyIFVq4Snv_WYvD3cv94-FYuXx-fbm0Wha9GNBVgmdK1Fw_uuarVRFkAwYXXTc7BYWz03NVaiF7yD3jSitxaRzdG0rWlaVR2Tiyn3Q61kXntQ8VsG5eTTzULu7hjUIICLLWT2fGLz0p8bTKMcXNK4WimPYZNk5homuqqrMlpPqI4hpYj2LxuY3AmXSzkJlzvheYrMwnPb9dSG-ctbh1GmnTSNxsXsRZrg_g_4AaGdjBs</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>1816089393</pqid></control><display><type>article</type><title>Dynamic attention priors: a new and efficient concept for improving object detection</title><source>Elsevier ScienceDirect Journals</source><creator>Gepperth, Alexander R.T. ; Ortiz, Michael Garcia ; Sattarov, Egor ; Heisele, Bernd</creator><creatorcontrib>Gepperth, Alexander R.T. ; Ortiz, Michael Garcia ; Sattarov, Egor ; Heisele, Bernd</creatorcontrib><description>Recent psychophysical evidence in humans suggests that visual attention is a highly dynamic and predictive process involving precise models of object trajectories. We present a proof-of-concept that such predictive spatial attention can benefit a technical system solving a challenging visual object detection task. To this end, we introduce a Bayes-like integration of the so-called dynamic attention priors (DAPs) and dense detection likelihoods, which get enhanced at predicted object positions obtained by the extrapolation of trajectories.
Using annotated video sequences of pedestrians in a parking lot setting, we quantitatively show that DAPs can improve detection performance significantly as compared to a baseline condition relying purely on pattern analysis.</description><identifier>ISSN: 0925-2312</identifier><identifier>EISSN: 1872-8286</identifier><identifier>DOI: 10.1016/j.neucom.2016.01.036</identifier><language>eng</language><publisher>Elsevier B.V</publisher><subject>Computer Science ; Dynamical systems ; Dynamics ; Extrapolation ; Machine Learning ; Mathematical models ; Object detection ; Pedestrians ; Tasks ; Trajectories ; Visual ; Visual attention</subject><ispartof>Neurocomputing (Amsterdam), 2016-07, Vol.197, p.14-28</ispartof><rights>2016 Elsevier B.V.</rights><rights>Distributed under a Creative Commons Attribution 4.0 International License</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c489t-1f08c4c862b937cdaf11808fc6b21fe4fc5d4e38b8291bd68bffee05ed77d67a3</citedby><cites>FETCH-LOGICAL-c489t-1f08c4c862b937cdaf11808fc6b21fe4fc5d4e38b8291bd68bffee05ed77d67a3</cites><orcidid>0000-0001-8232-5156</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://www.sciencedirect.com/science/article/pii/S0925231216001259$$EHTML$$P50$$Gelsevier$$H</linktohtml><link.rule.ids>230,314,776,780,881,3537,27901,27902,65306</link.rule.ids><backlink>$$Uhttps://hal.science/hal-01418128$$DView record in HAL$$Hfree_for_read</backlink></links><search><creatorcontrib>Gepperth, Alexander R.T.</creatorcontrib><creatorcontrib>Ortiz, Michael Garcia</creatorcontrib><creatorcontrib>Sattarov, Egor</creatorcontrib><creatorcontrib>Heisele, Bernd</creatorcontrib><title>Dynamic attention priors: a new and efficient concept for improving object detection</title><title>Neurocomputing (Amsterdam)</title><description>Recent psychophysical evidence in humans suggests that visual attention is a highly dynamic and predictive process involving precise models of object trajectories. We present a proof-of-concept that such predictive spatial attention can benefit a technical system solving a challenging visual object detection task. To this end, we introduce a Bayes-like integration of the so-called dynamic attention priors (DAPs) and dense detection likelihoods, which get enhanced at predicted object positions obtained by the extrapolation of trajectories.
Using annotated video sequences of pedestrians in a parking lot setting, we quantitatively show that DAPs can improve detection performance significantly as compared to a baseline condition relying purely on pattern analysis.</description><subject>Computer Science</subject><subject>Dynamical systems</subject><subject>Dynamics</subject><subject>Extrapolation</subject><subject>Machine Learning</subject><subject>Mathematical models</subject><subject>Object detection</subject><subject>Pedestrians</subject><subject>Tasks</subject><subject>Trajectories</subject><subject>Visual</subject><subject>Visual attention</subject><issn>0925-2312</issn><issn>1872-8286</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2016</creationdate><recordtype>article</recordtype><recordid>eNp9kE1PxCAQhonRxPXjH3jgqIdWhnZb6sHE-J1s4kXPhMKgbLawQneN_142NR49TYBn3hkeQs6AlcCguVyWHjc6DCXPp5JByapmj8xAtLwQXDT7ZMY6Pi94BfyQHKW0ZAxa4N2MvN59ezU4TdU4oh9d8HQdXYjpiirq8Ysqbyha67TLz1QHr3E9UhsidcM6hq3z7zT0S9QjNTjmkiNOyIFVq4Snv_WYvD3cv94-FYuXx-fbm0Wha9GNBVgmdK1Fw_uuarVRFkAwYXXTc7BYWz03NVaiF7yD3jSitxaRzdG0rWlaVR2Tiyn3Q61kXntQ8VsG5eTTzULu7hjUIICLLWT2fGLz0p8bTKMcXNK4WimPYZNk5homuqqrMlpPqI4hpYj2LxuY3AmXSzkJlzvheYrMwnPb9dSG-ctbh1GmnTSNxsXsRZrg_g_4AaGdjBs</recordid><startdate>20160712</startdate><enddate>20160712</enddate><creator>Gepperth, Alexander R.T.</creator><creator>Ortiz, Michael Garcia</creator><creator>Sattarov, Egor</creator><creator>Heisele, Bernd</creator><general>Elsevier B.V</general><general>Elsevier</general><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope><scope>1XC</scope><scope>VOOES</scope><orcidid>https://orcid.org/0000-0001-8232-5156</orcidid></search><sort><creationdate>20160712</creationdate><title>Dynamic attention priors: a new and efficient concept for improving object detection</title><author>Gepperth, Alexander R.T. ; Ortiz, Michael Garcia ; Sattarov, Egor ; Heisele, Bernd</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c489t-1f08c4c862b937cdaf11808fc6b21fe4fc5d4e38b8291bd68bffee05ed77d67a3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2016</creationdate><topic>Computer Science</topic><topic>Dynamical systems</topic><topic>Dynamics</topic><topic>Extrapolation</topic><topic>Machine Learning</topic><topic>Mathematical models</topic><topic>Object detection</topic><topic>Pedestrians</topic><topic>Tasks</topic><topic>Trajectories</topic><topic>Visual</topic><topic>Visual attention</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Gepperth, Alexander R.T.</creatorcontrib><creatorcontrib>Ortiz, Michael Garcia</creatorcontrib><creatorcontrib>Sattarov, Egor</creatorcontrib><creatorcontrib>Heisele, Bernd</creatorcontrib><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Technology Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><collection>Hyper Article en Ligne (HAL)</collection><collection>Hyper Article en Ligne (HAL) (Open Access)</collection><jtitle>Neurocomputing (Amsterdam)</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Gepperth, Alexander R.T.</au><au>Ortiz, Michael Garcia</au><au>Sattarov, Egor</au><au>Heisele, Bernd</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Dynamic attention priors: a new and efficient concept for improving object detection</atitle><jtitle>Neurocomputing (Amsterdam)</jtitle><date>2016-07-12</date><risdate>2016</risdate><volume>197</volume><spage>14</spage><epage>28</epage><pages>14-28</pages><issn>0925-2312</issn><eissn>1872-8286</eissn><abstract>Recent psychophysical evidence in humans suggests that visual attention is a highly dynamic and predictive process involving precise models of object trajectories. We present a proof-of-concept that such predictive spatial attention can benefit a technical system solving a challenging visual object detection task. To this end, we introduce a Bayes-like integration of the so-called dynamic attention priors (DAPs) and dense detection likelihoods, which get enhanced at predicted object positions obtained by the extrapolation of trajectories.
Using annotated video sequences of pedestrians in a parking lot setting, we quantitatively show that DAPs can improve detection performance significantly as compared to a baseline condition relying purely on pattern analysis.</abstract><pub>Elsevier B.V</pub><doi>10.1016/j.neucom.2016.01.036</doi><tpages>15</tpages><orcidid>https://orcid.org/0000-0001-8232-5156</orcidid><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext |
| identifier | ISSN: 0925-2312 |
| ispartof | Neurocomputing (Amsterdam), 2016-07, Vol.197, p.14-28 |
| issn | 0925-2312 1872-8286 |
| language | eng |
| recordid | cdi_hal_primary_oai_HAL_hal_01418128v1 |
| source | Elsevier ScienceDirect Journals |
| subjects | Computer Science Dynamical systems Dynamics Extrapolation Machine Learning Mathematical models Object detection Pedestrians Tasks Trajectories Visual Visual attention |
| title | Dynamic attention priors: a new and efficient concept for improving object detection |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-03T17%3A37%3A37IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_hal_p&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Dynamic%20attention%20priors:%20a%20new%20and%20efficient%20concept%20for%20improving%20object%20detection&rft.jtitle=Neurocomputing%20(Amsterdam)&rft.au=Gepperth,%20Alexander%20R.T.&rft.date=2016-07-12&rft.volume=197&rft.spage=14&rft.epage=28&rft.pages=14-28&rft.issn=0925-2312&rft.eissn=1872-8286&rft_id=info:doi/10.1016/j.neucom.2016.01.036&rft_dat=%3Cproquest_hal_p%3E1816089393%3C/proquest_hal_p%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=1816089393&rft_id=info:pmid/&rft_els_id=S0925231216001259&rfr_iscdi=true |