A Study on Quantifying Sim2Real Image Gap in Autonomous Driving Simulations Using Lane Segmentation Attention Map Similarity
Autonomous driving simulations require highly realistic images. Our preliminary study found that when the CARLA Simulator image was made more like reality by using DCLGAN, the performance of the lane recognition model improved to levels comparable to real-world driving. It was also confirmed that th...
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Artikel |
| Sprache: | 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 | Park, Seongjeong Pahk, Jinu Jahn, Lennart Lorenz Freimuth Lim, Yongseob An, Jinung Choi, Gyeungho |
| description | Autonomous driving simulations require highly realistic images. Our
preliminary study found that when the CARLA Simulator image was made more like
reality by using DCLGAN, the performance of the lane recognition model improved
to levels comparable to real-world driving. It was also confirmed that the
vehicle's ability to return to the center of the lane after deviating from it
improved significantly. However, there is currently no agreed-upon metric for
quantitatively evaluating the realism of simulation images. To address this
issue, based on the idea that FID (Fr\'echet Inception Distance) measures the
feature vector distribution distance using a pre-trained model, this paper
proposes a metric that measures the similarity of simulation road images using
the attention map from the self-attention distillation process of ENet-SAD.
Finally, this paper verified the suitability of the measurement method by
applying it to the image of the CARLA map that implemented a realworld
autonomous driving test road. |
| doi_str_mv | 10.48550/arxiv.2306.10491 |
| format | Article |
| fullrecord | <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2306_10491</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2306_10491</sourcerecordid><originalsourceid>FETCH-LOGICAL-a671-bb1cca1ec594118b915dcf69f27146f64fa6c079c4422d0f79e5ed96ff3184dc3</originalsourceid><addsrcrecordid>eNotkEFPhDAQhbl4MKs_wJPzB8AWSqFHsuq6CcYo65kMpSVNoGygbCTxxwu4p_ky781L5nneAyUBS-OYPOHwYy5BGBEeUMIEvfV-MyjcVM_QW_ic0DqjZ2MbKEwXfils4dhho-CAZzAWssn1tu_6aYTnwVyuxqlFZ3o7wve4bnK0CgrVdMq6TYDMuYVXel9ylgvT4mDcfOfdaGxHdX-dO-_0-nLav_n5x-G4z3IfeUL9qqJSIlUyFozStBI0rqXmQocJZVxzppFLkgjJWBjWRCdCxaoWXOuIpqyW0c57_I_d3i_Pg-lwmMu1hnKrIfoDXWVZ8w</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>A Study on Quantifying Sim2Real Image Gap in Autonomous Driving Simulations Using Lane Segmentation Attention Map Similarity</title><source>arXiv.org</source><creator>Park, Seongjeong ; Pahk, Jinu ; Jahn, Lennart Lorenz Freimuth ; Lim, Yongseob ; An, Jinung ; Choi, Gyeungho</creator><creatorcontrib>Park, Seongjeong ; Pahk, Jinu ; Jahn, Lennart Lorenz Freimuth ; Lim, Yongseob ; An, Jinung ; Choi, Gyeungho</creatorcontrib><description>Autonomous driving simulations require highly realistic images. Our
preliminary study found that when the CARLA Simulator image was made more like
reality by using DCLGAN, the performance of the lane recognition model improved
to levels comparable to real-world driving. It was also confirmed that the
vehicle's ability to return to the center of the lane after deviating from it
improved significantly. However, there is currently no agreed-upon metric for
quantitatively evaluating the realism of simulation images. To address this
issue, based on the idea that FID (Fr\'echet Inception Distance) measures the
feature vector distribution distance using a pre-trained model, this paper
proposes a metric that measures the similarity of simulation road images using
the attention map from the self-attention distillation process of ENet-SAD.
Finally, this paper verified the suitability of the measurement method by
applying it to the image of the CARLA map that implemented a realworld
autonomous driving test road.</description><identifier>DOI: 10.48550/arxiv.2306.10491</identifier><language>eng</language><subject>Computer Science - Computer Vision and Pattern Recognition ; Computer Science - Robotics</subject><creationdate>2023-06</creationdate><rights>http://creativecommons.org/licenses/by/4.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,780,885</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2306.10491$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2306.10491$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Park, Seongjeong</creatorcontrib><creatorcontrib>Pahk, Jinu</creatorcontrib><creatorcontrib>Jahn, Lennart Lorenz Freimuth</creatorcontrib><creatorcontrib>Lim, Yongseob</creatorcontrib><creatorcontrib>An, Jinung</creatorcontrib><creatorcontrib>Choi, Gyeungho</creatorcontrib><title>A Study on Quantifying Sim2Real Image Gap in Autonomous Driving Simulations Using Lane Segmentation Attention Map Similarity</title><description>Autonomous driving simulations require highly realistic images. Our
preliminary study found that when the CARLA Simulator image was made more like
reality by using DCLGAN, the performance of the lane recognition model improved
to levels comparable to real-world driving. It was also confirmed that the
vehicle's ability to return to the center of the lane after deviating from it
improved significantly. However, there is currently no agreed-upon metric for
quantitatively evaluating the realism of simulation images. To address this
issue, based on the idea that FID (Fr\'echet Inception Distance) measures the
feature vector distribution distance using a pre-trained model, this paper
proposes a metric that measures the similarity of simulation road images using
the attention map from the self-attention distillation process of ENet-SAD.
Finally, this paper verified the suitability of the measurement method by
applying it to the image of the CARLA map that implemented a realworld
autonomous driving test road.</description><subject>Computer Science - Computer Vision and Pattern Recognition</subject><subject>Computer Science - Robotics</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotkEFPhDAQhbl4MKs_wJPzB8AWSqFHsuq6CcYo65kMpSVNoGygbCTxxwu4p_ky781L5nneAyUBS-OYPOHwYy5BGBEeUMIEvfV-MyjcVM_QW_ic0DqjZ2MbKEwXfils4dhho-CAZzAWssn1tu_6aYTnwVyuxqlFZ3o7wve4bnK0CgrVdMq6TYDMuYVXel9ylgvT4mDcfOfdaGxHdX-dO-_0-nLav_n5x-G4z3IfeUL9qqJSIlUyFozStBI0rqXmQocJZVxzppFLkgjJWBjWRCdCxaoWXOuIpqyW0c57_I_d3i_Pg-lwmMu1hnKrIfoDXWVZ8w</recordid><startdate>20230618</startdate><enddate>20230618</enddate><creator>Park, Seongjeong</creator><creator>Pahk, Jinu</creator><creator>Jahn, Lennart Lorenz Freimuth</creator><creator>Lim, Yongseob</creator><creator>An, Jinung</creator><creator>Choi, Gyeungho</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20230618</creationdate><title>A Study on Quantifying Sim2Real Image Gap in Autonomous Driving Simulations Using Lane Segmentation Attention Map Similarity</title><author>Park, Seongjeong ; Pahk, Jinu ; Jahn, Lennart Lorenz Freimuth ; Lim, Yongseob ; An, Jinung ; Choi, Gyeungho</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a671-bb1cca1ec594118b915dcf69f27146f64fa6c079c4422d0f79e5ed96ff3184dc3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Computer Science - Computer Vision and Pattern Recognition</topic><topic>Computer Science - Robotics</topic><toplevel>online_resources</toplevel><creatorcontrib>Park, Seongjeong</creatorcontrib><creatorcontrib>Pahk, Jinu</creatorcontrib><creatorcontrib>Jahn, Lennart Lorenz Freimuth</creatorcontrib><creatorcontrib>Lim, Yongseob</creatorcontrib><creatorcontrib>An, Jinung</creatorcontrib><creatorcontrib>Choi, Gyeungho</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Park, Seongjeong</au><au>Pahk, Jinu</au><au>Jahn, Lennart Lorenz Freimuth</au><au>Lim, Yongseob</au><au>An, Jinung</au><au>Choi, Gyeungho</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>A Study on Quantifying Sim2Real Image Gap in Autonomous Driving Simulations Using Lane Segmentation Attention Map Similarity</atitle><date>2023-06-18</date><risdate>2023</risdate><abstract>Autonomous driving simulations require highly realistic images. Our
preliminary study found that when the CARLA Simulator image was made more like
reality by using DCLGAN, the performance of the lane recognition model improved
to levels comparable to real-world driving. It was also confirmed that the
vehicle's ability to return to the center of the lane after deviating from it
improved significantly. However, there is currently no agreed-upon metric for
quantitatively evaluating the realism of simulation images. To address this
issue, based on the idea that FID (Fr\'echet Inception Distance) measures the
feature vector distribution distance using a pre-trained model, this paper
proposes a metric that measures the similarity of simulation road images using
the attention map from the self-attention distillation process of ENet-SAD.
Finally, this paper verified the suitability of the measurement method by
applying it to the image of the CARLA map that implemented a realworld
autonomous driving test road.</abstract><doi>10.48550/arxiv.2306.10491</doi><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | DOI: 10.48550/arxiv.2306.10491 |
| ispartof | |
| issn | |
| language | eng |
| recordid | cdi_arxiv_primary_2306_10491 |
| source | arXiv.org |
| subjects | Computer Science - Computer Vision and Pattern Recognition Computer Science - Robotics |
| title | A Study on Quantifying Sim2Real Image Gap in Autonomous Driving Simulations Using Lane Segmentation Attention Map Similarity |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-27T09%3A05%3A23IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=A%20Study%20on%20Quantifying%20Sim2Real%20Image%20Gap%20in%20Autonomous%20Driving%20Simulations%20Using%20Lane%20Segmentation%20Attention%20Map%20Similarity&rft.au=Park,%20Seongjeong&rft.date=2023-06-18&rft_id=info:doi/10.48550/arxiv.2306.10491&rft_dat=%3Carxiv_GOX%3E2306_10491%3C/arxiv_GOX%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 |