Dangerous driving behavior identification method based on sample imbalance
The invention discloses a dangerous driving behavior identification method based on sample imbalance. The method comprises the following steps: (1) performing data enhancement processing on collected few-sample category data; (2) carrying out oversampling processing on the few-sample category data s...
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Patent |
| Sprache: | chi ; eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | |
|---|---|
| container_issue | |
| container_start_page | |
| container_title | |
| container_volume | |
| creator | HU KAIHUA ZHOU DESONG XU GUILIN WU SHUAIXIAN |
| description | The invention discloses a dangerous driving behavior identification method based on sample imbalance. The method comprises the following steps: (1) performing data enhancement processing on collected few-sample category data; (2) carrying out oversampling processing on the few-sample category data subjected to data enhancement processing; (3) adopting the OfficientNet-B0 as a baseline model, and using a cross entropy loss function WCE as a baseline loss function to carry out weight equalization on the data of the few sample categories; (4) carrying out regularization processing on the cross entropy loss function WCE through an L2 norm; and (5) a convolution attention module (CBAM) is added to the OfficientNet-B0 baseline model. According to the invention, the oversampling regular strategy model is provided to realize the optimization design of the dangerous driving behavior recognition model under the condition of sample imbalance, the detection rate and accuracy of the dangerous driving behavior are greatly |
| format | Patent |
| fullrecord | <record><control><sourceid>epo_EVB</sourceid><recordid>TN_cdi_epo_espacenet_CN117809290A</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>CN117809290A</sourcerecordid><originalsourceid>FETCH-epo_espacenet_CN117809290A3</originalsourceid><addsrcrecordid>eNqNyjsKwkAURuE0FqLu4boAIdFCU0pUxMLKPtyZ-ZNcmBczY9avhQuwOnxwltXjwn5ECu9MJsksfiSFiWcJicTAFxlEc5HgyaFMwZDiDENfZ3bRgsQptuw11tViYJux-XVVbW_XV3ffIYYeObKGR-m7Z9McT3W7b-vz4Z_nA1PPNWs</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>patent</recordtype></control><display><type>patent</type><title>Dangerous driving behavior identification method based on sample imbalance</title><source>esp@cenet</source><creator>HU KAIHUA ; ZHOU DESONG ; XU GUILIN ; WU SHUAIXIAN</creator><creatorcontrib>HU KAIHUA ; ZHOU DESONG ; XU GUILIN ; WU SHUAIXIAN</creatorcontrib><description>The invention discloses a dangerous driving behavior identification method based on sample imbalance. The method comprises the following steps: (1) performing data enhancement processing on collected few-sample category data; (2) carrying out oversampling processing on the few-sample category data subjected to data enhancement processing; (3) adopting the OfficientNet-B0 as a baseline model, and using a cross entropy loss function WCE as a baseline loss function to carry out weight equalization on the data of the few sample categories; (4) carrying out regularization processing on the cross entropy loss function WCE through an L2 norm; and (5) a convolution attention module (CBAM) is added to the OfficientNet-B0 baseline model. According to the invention, the oversampling regular strategy model is provided to realize the optimization design of the dangerous driving behavior recognition model under the condition of sample imbalance, the detection rate and accuracy of the dangerous driving behavior are greatly</description><language>chi ; eng</language><subject>ALARM SYSTEMS ; CALCULATING ; COMPUTING ; COUNTING ; ORDER TELEGRAPHS ; PHYSICS ; SIGNALLING ; SIGNALLING OR CALLING SYSTEMS</subject><creationdate>2024</creationdate><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20240402&DB=EPODOC&CC=CN&NR=117809290A$$EHTML$$P50$$Gepo$$Hfree_for_read</linktohtml><link.rule.ids>230,308,776,881,25543,76293</link.rule.ids><linktorsrc>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20240402&DB=EPODOC&CC=CN&NR=117809290A$$EView_record_in_European_Patent_Office$$FView_record_in_$$GEuropean_Patent_Office$$Hfree_for_read</linktorsrc></links><search><creatorcontrib>HU KAIHUA</creatorcontrib><creatorcontrib>ZHOU DESONG</creatorcontrib><creatorcontrib>XU GUILIN</creatorcontrib><creatorcontrib>WU SHUAIXIAN</creatorcontrib><title>Dangerous driving behavior identification method based on sample imbalance</title><description>The invention discloses a dangerous driving behavior identification method based on sample imbalance. The method comprises the following steps: (1) performing data enhancement processing on collected few-sample category data; (2) carrying out oversampling processing on the few-sample category data subjected to data enhancement processing; (3) adopting the OfficientNet-B0 as a baseline model, and using a cross entropy loss function WCE as a baseline loss function to carry out weight equalization on the data of the few sample categories; (4) carrying out regularization processing on the cross entropy loss function WCE through an L2 norm; and (5) a convolution attention module (CBAM) is added to the OfficientNet-B0 baseline model. According to the invention, the oversampling regular strategy model is provided to realize the optimization design of the dangerous driving behavior recognition model under the condition of sample imbalance, the detection rate and accuracy of the dangerous driving behavior are greatly</description><subject>ALARM SYSTEMS</subject><subject>CALCULATING</subject><subject>COMPUTING</subject><subject>COUNTING</subject><subject>ORDER TELEGRAPHS</subject><subject>PHYSICS</subject><subject>SIGNALLING</subject><subject>SIGNALLING OR CALLING SYSTEMS</subject><fulltext>true</fulltext><rsrctype>patent</rsrctype><creationdate>2024</creationdate><recordtype>patent</recordtype><sourceid>EVB</sourceid><recordid>eNqNyjsKwkAURuE0FqLu4boAIdFCU0pUxMLKPtyZ-ZNcmBczY9avhQuwOnxwltXjwn5ECu9MJsksfiSFiWcJicTAFxlEc5HgyaFMwZDiDENfZ3bRgsQptuw11tViYJux-XVVbW_XV3ffIYYeObKGR-m7Z9McT3W7b-vz4Z_nA1PPNWs</recordid><startdate>20240402</startdate><enddate>20240402</enddate><creator>HU KAIHUA</creator><creator>ZHOU DESONG</creator><creator>XU GUILIN</creator><creator>WU SHUAIXIAN</creator><scope>EVB</scope></search><sort><creationdate>20240402</creationdate><title>Dangerous driving behavior identification method based on sample imbalance</title><author>HU KAIHUA ; ZHOU DESONG ; XU GUILIN ; WU SHUAIXIAN</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-epo_espacenet_CN117809290A3</frbrgroupid><rsrctype>patents</rsrctype><prefilter>patents</prefilter><language>chi ; eng</language><creationdate>2024</creationdate><topic>ALARM SYSTEMS</topic><topic>CALCULATING</topic><topic>COMPUTING</topic><topic>COUNTING</topic><topic>ORDER TELEGRAPHS</topic><topic>PHYSICS</topic><topic>SIGNALLING</topic><topic>SIGNALLING OR CALLING SYSTEMS</topic><toplevel>online_resources</toplevel><creatorcontrib>HU KAIHUA</creatorcontrib><creatorcontrib>ZHOU DESONG</creatorcontrib><creatorcontrib>XU GUILIN</creatorcontrib><creatorcontrib>WU SHUAIXIAN</creatorcontrib><collection>esp@cenet</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>HU KAIHUA</au><au>ZHOU DESONG</au><au>XU GUILIN</au><au>WU SHUAIXIAN</au><format>patent</format><genre>patent</genre><ristype>GEN</ristype><title>Dangerous driving behavior identification method based on sample imbalance</title><date>2024-04-02</date><risdate>2024</risdate><abstract>The invention discloses a dangerous driving behavior identification method based on sample imbalance. The method comprises the following steps: (1) performing data enhancement processing on collected few-sample category data; (2) carrying out oversampling processing on the few-sample category data subjected to data enhancement processing; (3) adopting the OfficientNet-B0 as a baseline model, and using a cross entropy loss function WCE as a baseline loss function to carry out weight equalization on the data of the few sample categories; (4) carrying out regularization processing on the cross entropy loss function WCE through an L2 norm; and (5) a convolution attention module (CBAM) is added to the OfficientNet-B0 baseline model. According to the invention, the oversampling regular strategy model is provided to realize the optimization design of the dangerous driving behavior recognition model under the condition of sample imbalance, the detection rate and accuracy of the dangerous driving behavior are greatly</abstract><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | |
| ispartof | |
| issn | |
| language | chi ; eng |
| recordid | cdi_epo_espacenet_CN117809290A |
| source | esp@cenet |
| subjects | ALARM SYSTEMS CALCULATING COMPUTING COUNTING ORDER TELEGRAPHS PHYSICS SIGNALLING SIGNALLING OR CALLING SYSTEMS |
| title | Dangerous driving behavior identification method based on sample imbalance |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-24T15%3A37%3A48IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-epo_EVB&rft_val_fmt=info:ofi/fmt:kev:mtx:patent&rft.genre=patent&rft.au=HU%20KAIHUA&rft.date=2024-04-02&rft_id=info:doi/&rft_dat=%3Cepo_EVB%3ECN117809290A%3C/epo_EVB%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true |