As the ACCIDENT challenge has ended, we would like to invite participants to submit their work to the AUTOPILOT Workshop.
\nTake this as a good opportunity to develop your challenge submission into a workshop paper and present it, as selected papers will have the opportunity to be presented at CVPR through the AUTOPILOT Workshop. In addition to reporting final performance, a submission should highlight the main idea behind your approach, design choices, ablations, error analysis, or practical lessons learned during the competition.
\nWe especially encourage submissions that go beyond leaderboard ranking alone and help others better understand what worked, what did not, and which directions seem promising on the ACCIDENT benchmark.
\nTo support follow-up work, we have been working hard on several things:
\nP.S. We are still actively working on the website and related resources. Some parts may still be incomplete or temporarily unavailable as we continue to verify everything and finalize the setup. Please do not use the results returned by the website YET!
", "messages": [ { "id": "3441401", "postDate": "04/13/2026 23:09:35", "content": "As the ACCIDENT challenge has ended, we would like to invite participants to submit their work to the AUTOPILOT Workshop.
\nTake this as a good opportunity to develop your challenge submission into a workshop paper and present it, as selected papers will have the opportunity to be presented at CVPR through the AUTOPILOT Workshop. In addition to reporting final performance, a submission should highlight the main idea behind your approach, design choices, ablations, error analysis, or practical lessons learned during the competition.
\nWe especially encourage submissions that go beyond leaderboard ranking alone and help others better understand what worked, what did not, and which directions seem promising on the ACCIDENT benchmark.
\nTo support follow-up work, we have been working hard on several things:
\nP.S. We are still actively working on the website and related resources. Some parts may still be incomplete or temporarily unavailable as we continue to verify everything and finalize the setup. Please do not use the results returned by the website YET!
", "rawMarkdown": "As the ACCIDENT challenge has ended, we would like to invite participants to submit their work to the [AUTOPILOT Workshop](https://www.autopilot-cvpr.net/).\n\nTake this as a good opportunity to develop your challenge submission into a workshop paper and present it, as **selected papers will have the opportunity to be presented at CVPR through the AUTOPILOT Workshop.** In addition to reporting final performance, a submission should highlight the main idea behind your approach, design choices, ablations, error analysis, or practical lessons learned during the competition.\n\nWe especially encourage submissions that go beyond leaderboard ranking alone and help others better understand what worked, what did not, and which directions seem promising on the ACCIDENT benchmark.\n\nTo support follow-up work, we have been working hard on several things:\n- Benchmark website with public leaderboards, submission instructions, and a browser-based scoring interface for CSV submissions --> https://accidentbench.github.io/\n- ACCIDENT dataset with labels available on Kaggle --> https://www.kaggle.com/datasets/picekl/accident\n- Dataset paper with full technical details, benchmark design, and evaluation methodology --> https://arxiv.org/abs/2604.09819\n\n**P.S. We are still actively working on the website and related resources. Some parts may still be incomplete or temporarily unavailable as we continue to verify everything and finalize the setup. Please do not use the results returned by the website YET!**", "votes": null }, { "id": "3441402", "postDate": "04/13/2026 23:12:22", "content": "@aadigupta1601, @abderrahmen36, @abhaylejith, @adibraian, @adityasiraskar, @ahnaftahmid24, @aira666, @aisang, @aiswarjya, @akumar49, @ameythakur20, @andrewagirgis, @anidiptapal, @anonamename, @apuchakroborti, @arwaabouattia, @asmaasamih22, @asmaasamyh2020, @asmaasamyhagag22, @atomkevich, @barshamishra19, @belkessambadis, @bigshark101, @birunthabanrajendram, @brendayyc, @buharisalisu, @calpis10000, @chancekim, @chayakornwinuntamala, @chengsiqi47, @chenruwen, @choigu, @choisanghyuk, @chow123, @csanskriti, @dhruvpaidukle, @dionysisvoulgarakis, @dipit099, @dkwlsdk, @dla020501, @dqliu12138, @dubbin, @erdespiro, @fariawanpakistan12, @feel12348, @fengsiyi12345678, @fernandollsalvador, @ftakahas, @fuumin621, @gch8678, @gouthamk16, @guangjiandong, @hariprasanth8, @hepingpeace, @hiro356, @hmnshudhmn24, @hongyuchen123321, @hwanghwan, @hwkims, @hyobink, @icebearogo, @ihkali, @irajmoradi, @irisreto, @ishankat, @jayhawk1900, @jek1wantaufik, @jianthuang, @jinchengzhang1, @jooyoungseok, @joshuaburnsblack, @jsalvasoler, @kiyotaka1102, @ksingh51, @kwarkamole, @lakhindarpal, @letita, @lijiaowang97, @lollylandd, @lsiavchangpeng, @lunaandendymion, @mapjoo, @mathurinache, @maximshatskiy, @mdakramhossain1871, @mirzamilanfarabi, @mrpc2003, @ngocnhannguyen, @nhnnguynngc, @novebrititoramadhani, @oliverlang, @parkerke, @patrickiitmz, @patricktz, @peacelu, @pensukesan, @phanisrikanth, @prashanttandon, @ps014arkyabagchi, @raihan0001, @raihanrsd, @ratnachand, @rianryu, @ricswhat, @rituljangir, @romeanthem, @ruitongs, @ruwadnaswan, @rwdnaswan, @ryudonghun, @sachinkukkar, @sanjidh090, @saravananj, @sarveshtalele, @seongwonk, @seongwonkang, @shahadsawla, @shakhoyatshujon, @shouta0108, @shyngys19732, @sigmaborov, @simonartzet, @sirteo, @sjskmu, @skssdn, @skuldnorniern, @smak17, @smammahdi, @smammer056, @smcho1201, @snigdhatalasila, @soyaoki, @suam6409, @taejinpark0727, @takamaruex, @tarandeep97, @thiendangnn63, @tjrhyun, @tnr1337, @tonyironman099, @trikang, @varunraskar, @wenjiozhao, @xiaoqiaodong, @xinyvwang, @yashanathaniel, @yeonslay, @yoyy0099, @yuxiazff, @yuxisu944, @zahidbooni
", "rawMarkdown": "aadigupta1601, @abderrahmen36, @abhaylejith, @adibraian, @adityasiraskar, @ahnaftahmid24, @aira666, @aisang, @aiswarjya, @akumar49, @ameythakur20, @andrewagirgis, @anidiptapal, @anonamename, @apuchakroborti, @arwaabouattia, @asmaasamih22, @asmaasamyh2020, @asmaasamyhagag22, @atomkevich, @barshamishra19, @belkessambadis, @bigshark101, @birunthabanrajendram, @brendayyc, @buharisalisu, @calpis10000, @chancekim, @chayakornwinuntamala, @chengsiqi47, @chenruwen, @choigu, @choisanghyuk, @chow123, @csanskriti, @dhruvpaidukle, @dionysisvoulgarakis, @dipit099, @dkwlsdk, @dla020501, @dqliu12138, @dubbin, @erdespiro, @fariawanpakistan12, @feel12348, @fengsiyi12345678, @fernandollsalvador, @ftakahas, @fuumin621, @gch8678, @gouthamk16, @guangjiandong, @hariprasanth8, @hepingpeace, @hiro356, @hmnshudhmn24, @hongyuchen123321, @hwanghwan, @hwkims, @hyobink, @icebearogo, @ihkali, @irajmoradi, @irisreto, @ishankat, @jayhawk1900, @jek1wantaufik, @jianthuang, @jinchengzhang1, @jooyoungseok, @joshuaburnsblack, @jsalvasoler, @kiyotaka1102, @ksingh51, @kwarkamole, @lakhindarpal, @letita, @lijiaowang97, @lollylandd, @lsiavchangpeng, @lunaandendymion, @mapjoo, @mathurinache, @maximshatskiy, @mdakramhossain1871, @mirzamilanfarabi, @mrpc2003, @ngocnhannguyen, @nhnnguynngc, @novebrititoramadhani, @oliverlang, @parkerke, @patrickiitmz, @patricktz, @peacelu, @pensukesan, @phanisrikanth, @prashanttandon, @ps014arkyabagchi, @raihan0001, @raihanrsd, @ratnachand, @rianryu, @ricswhat, @rituljangir, @romeanthem, @ruitongs, @ruwadnaswan, @rwdnaswan, @ryudonghun, @sachinkukkar, @sanjidh090, @saravananj, @sarveshtalele, @seongwonk, @seongwonkang, @shahadsawla, @shakhoyatshujon, @shouta0108, @shyngys19732, @sigmaborov, @simonartzet, @sirteo, @sjskmu, @skssdn, @skuldnorniern, @smak17, @smammahdi, @smammer056, @smcho1201, @snigdhatalasila, @soyaoki, @suam6409, @taejinpark0727, @takamaruex, @tarandeep97, @thiendangnn63, @tjrhyun, @tnr1337, @tonyironman099, @trikang, @varunraskar, @wenjiozhao, @xiaoqiaodong, @xinyvwang, @yashanathaniel, @yeonslay, @yoyy0099, @yuxiazff, @yuxisu944, @zahidbooni", "votes": null }, { "id": "3442278", "postDate": "04/15/2026 08:02:07", "content": "@picekl \nHi, I’m from the winning team of the ACCIDENT challenge. I'm currently preparing a paper for the workshop, but since I'm based in Japan, it's difficult to attend in person. Will there be an option for a virtual or pre-recorded video presentation? Thanks!
", "rawMarkdown": "picekl \nHi, I’m from the winning team of the ACCIDENT challenge. I'm currently preparing a paper for the workshop, but since I'm based in Japan, it's difficult to attend in person. Will there be an option for a virtual or pre-recorded video presentation? Thanks!", "votes": null }, { "id": "3442395", "postDate": "04/15/2026 12:37:32", "content": "Hi @fuumin621 ,
\nCongratulations on winning the competition!
\nAbout your question. We are aware of the situation. Thus, we are currently discussing the Workshop format with other Workshop chairs.\nIn any case, virtual participation or prerecorded videos should not be a problem. FYI, I prefer a virtual participation.
\nBest,\nLukas
", "rawMarkdown": "Hi @fuumin621 ,\n\nCongratulations on winning the competition!\n\nAbout your question. We are aware of the situation. Thus, we are currently discussing the Workshop format with other Workshop chairs.\nIn any case, virtual participation or prerecorded videos should not be a problem. FYI, I prefer a virtual participation.\n\nBest,\nLukas", "votes": null }, { "id": "3442960", "postDate": "04/16/2026 01:53:46", "content": "Thank you for the clarification! I’m relieved to hear that. I would prefer to participate virtually as well, so I’ll wait for the final details. Thanks again!
", "rawMarkdown": "Thank you for the clarification! I’m relieved to hear that. I would prefer to participate virtually as well, so I’ll wait for the final details. Thanks again!", "votes": null }, { "id": "3444023", "postDate": "04/17/2026 06:21:57", "content": "@picekl Hello,I am a member of the team that ranked 2nd in the VQA track. \nI noticed that a \"Call for Papers\" has been posted in the ACCIDENT track discussion, but there hasn't been a similar announcement in the VQA track yet. Therefore, I would like to clarify:\n1.Requirement: Are participants in the VQA track also expected to submit a paper for the workshop?\n2.Format: As we have a very tight schedule currently, would it be acceptable if we submitted a concise technical report instead of a full workshop paper?\n3.Logistics: Could you please provide the submission portal/link and the official deadline for the VQA track?
", "rawMarkdown": "picekl Hello,I am a member of the team that ranked 2nd in the VQA track. \nI noticed that a \"Call for Papers\" has been posted in the ACCIDENT track discussion, but there hasn't been a similar announcement in the VQA track yet. Therefore, I would like to clarify:\n1.Requirement: Are participants in the VQA track also expected to submit a paper for the workshop?\n2.Format: As we have a very tight schedule currently, would it be acceptable if we submitted a concise technical report instead of a full workshop paper?\n3.Logistics: Could you please provide the submission portal/link and the official deadline for the VQA track?", "votes": null }, { "id": "3445535", "postDate": "04/20/2026 02:51:33", "content": "Hi @yuxiazff ,
\nI have no idea. You have to ask the organizers of that competition. \nI have nothing to do with it.
\nGood luck!
\nBest,\nLukas
", "rawMarkdown": "Hi @yuxiazff ,\n\nI have no idea. You have to ask the organizers of that competition. \nI have nothing to do with it. \n\nGood luck!\n\nBest,\nLukas", "votes": null }, { "id": "3446314", "postDate": "04/21/2026 11:51:05", "content": "Hi, @picekl
\nI noticed that the Molmo-7B results on the leaderboard differ from those reported in Table 5 of the paper (especially for S and ACCS), while the Naive and Best-from-all baselines seem consistent across both.I was wondering if this discrepancy is due to differences in the inference pipeline (e.g., temporal sampling, prompting), dataset versioning, or evaluation protocol between the paper and the leaderboard?
\nWould really appreciate any clarification. Thanks!
", "rawMarkdown": "Hi, @picekl \n\nI noticed that the Molmo-7B results on the leaderboard differ from those reported in Table 5 of the paper (especially for S and ACCS), while the Naive and Best-from-all baselines seem consistent across both.I was wondering if this discrepancy is due to differences in the inference pipeline (e.g., temporal sampling, prompting), dataset versioning, or evaluation protocol between the paper and the leaderboard?\n\nWould really appreciate any clarification. Thanks!", "votes": null }, { "id": "3446436", "postDate": "04/21/2026 16:02:11", "content": "Hi @guangjiandong ,
\nWhat do you mean by Malmo-7B results on the leaderboard? \nIf you mean This Leaderboard, it is most likely just an old artifact that I'll fix ASAP.
\nIn any case, the results in the paper are 100%; we double-checked them multiple times.
\nBest,\nLukas
", "rawMarkdown": "Hi @guangjiandong ,\n\nWhat do you mean by Malmo-7B results on the leaderboard? \nIf you mean [This Leaderboard](https://accidentbench.github.io/benchmarks/zero-shot/index.html), it is most likely just an old artifact that I'll fix ASAP.\n\nIn any case, the results in the paper are 100%; we double-checked them multiple times.\n\nBest,\nLukas", "votes": null }, { "id": "3449481", "postDate": "04/28/2026 10:20:16", "content": "@picekl We are trying to make a submission to AUTOPILOT @ CVPR 2026 — Archival Track, but not getting an option to submit it like it's in the Non-Archival Track.
", "rawMarkdown": "picekl We are trying to make a submission to AUTOPILOT @ CVPR 2026 — Archival Track, but not getting an option to submit it like it's in the Non-Archival Track.", "votes": null }, { "id": "3449524", "postDate": "04/28/2026 12:53:47", "content": "Hi @varunraskar ,
\nNot sure what you mean.\nDoes this link not work for you?\nhttps://openreview.net/group?id=thecvf.com/CVPR/2026/Workshop/AUTOPILOT-NA&referrer=%5BHomepage%5D(%2F)#tab-your-consoles
\nBest,\nLukas
", "rawMarkdown": "Hi @varunraskar ,\n\nNot sure what you mean.\nDoes this link not work for you?\nhttps://openreview.net/group?id=thecvf.com/CVPR/2026/Workshop/AUTOPILOT-NA&referrer=%5BHomepage%5D(%2F)#tab-your-consoles\n\nBest,\nLukas", "votes": null }, { "id": "3449540", "postDate": "04/28/2026 13:51:17", "content": "Got it, Thank you
", "rawMarkdown": "Got it, Thank you", "votes": null }, { "id": "3456458", "postDate": "05/11/2026 23:39:40", "content": "@picekl Hi, just wanted to follow up on the Non-Archival Track results. The notification was scheduled for May 10th, but I haven't received any updates yet. With the Virtual Materials deadline approaching on May 20th, could you let us know when we might expect the results? Thank you!
", "rawMarkdown": "picekl Hi, just wanted to follow up on the Non-Archival Track results. The notification was scheduled for May 10th, but I haven't received any updates yet. With the Virtual Materials deadline approaching on May 20th, could you let us know when we might expect the results? Thank you!", "votes": null }, { "id": "3456464", "postDate": "05/12/2026 00:09:00", "content": "Hi @ryudonghun ,
\nI just double checked with the other competition organizers and we will send the feedback and notification of acceptance / rejection tomorrow. There are some small issues that caused delay. Sorry for that.
\nBest,\nLukas
", "rawMarkdown": "Hi @ryudonghun ,\n\nI just double checked with the other competition organizers and we will send the feedback and notification of acceptance / rejection tomorrow. There are some small issues that caused delay. Sorry for that.\n\nBest,\nLukas", "votes": null }, { "id": "3456476", "postDate": "05/12/2026 00:55:36", "content": "No worries at all! Thanks for letting me know. Looking forward to it!
", "rawMarkdown": "No worries at all! Thanks for letting me know. Looking forward to it!", "votes": null }, { "id": "3457614", "postDate": "05/14/2026 02:14:19", "content": "Hi @picekl \nSorry to bother you again with this! I completely understand that the team is busy and things can get delayed. I just wanted to check if there's any update on the notifications, as I really need to start making plans and preparing for my conference participation.\nThank you again for your time and hard work!
", "rawMarkdown": "Hi @picekl \nSorry to bother you again with this! I completely understand that the team is busy and things can get delayed. I just wanted to check if there's any update on the notifications, as I really need to start making plans and preparing for my conference participation.\nThank you again for your time and hard work!", "votes": null }, { "id": "3457894", "postDate": "05/14/2026 15:12:17", "content": "Hi @ryudonghun ,
\nI have provided inputs to the program chairs on Friday. I think they are waiting for feedback from other competition organizers so they can send everything in one batch.
\nI forwarded your concern, but sadly, there is nothing else I can do at the moment. 😔
\nBest,\nLukas
", "rawMarkdown": "Hi @ryudonghun ,\n\nI have provided inputs to the program chairs on Friday. I think they are waiting for feedback from other competition organizers so they can send everything in one batch. \n\nI forwarded your concern, but sadly, there is nothing else I can do at the moment. 😔\n\nBest,\nLukas", "votes": null }, { "id": "3457898", "postDate": "05/14/2026 15:22:51", "content": "Thank you so much for the transparent update and for advocating on my behalf. I completely understand that it is out of your hands now and that the program chairs need to coordinate everything. I will patiently wait for the official notification. Have a great weekend!
", "rawMarkdown": "Thank you so much for the transparent update and for advocating on my behalf. I completely understand that it is out of your hands now and that the program chairs need to coordinate everything. I will patiently wait for the official notification. Have a great weekend!", "votes": null } ], "comments": [ { "id": 3441402, "author_name": "picekl", "author_url": "", "post_date": "04/13/2026 23:12:22", "content": "@aadigupta1601, @abderrahmen36, @abhaylejith, @adibraian, @adityasiraskar, @ahnaftahmid24, @aira666, @aisang, @aiswarjya, @akumar49, @ameythakur20, @andrewagirgis, @anidiptapal, @anonamename, @apuchakroborti, @arwaabouattia, @asmaasamih22, @asmaasamyh2020, @asmaasamyhagag22, @atomkevich, @barshamishra19, @belkessambadis, @bigshark101, @birunthabanrajendram, @brendayyc, @buharisalisu, @calpis10000, @chancekim, @chayakornwinuntamala, @chengsiqi47, @chenruwen, @choigu, @choisanghyuk, @chow123, @csanskriti, @dhruvpaidukle, @dionysisvoulgarakis, @dipit099, @dkwlsdk, @dla020501, @dqliu12138, @dubbin, @erdespiro, @fariawanpakistan12, @feel12348, @fengsiyi12345678, @fernandollsalvador, @ftakahas, @fuumin621, @gch8678, @gouthamk16, @guangjiandong, @hariprasanth8, @hepingpeace, @hiro356, @hmnshudhmn24, @hongyuchen123321, @hwanghwan, @hwkims, @hyobink, @icebearogo, @ihkali, @irajmoradi, @irisreto, @ishankat, @jayhawk1900, @jek1wantaufik, @jianthuang, @jinchengzhang1, @jooyoungseok, @joshuaburnsblack, @jsalvasoler, @kiyotaka1102, @ksingh51, @kwarkamole, @lakhindarpal, @letita, @lijiaowang97, @lollylandd, @lsiavchangpeng, @lunaandendymion, @mapjoo, @mathurinache, @maximshatskiy, @mdakramhossain1871, @mirzamilanfarabi, @mrpc2003, @ngocnhannguyen, @nhnnguynngc, @novebrititoramadhani, @oliverlang, @parkerke, @patrickiitmz, @patricktz, @peacelu, @pensukesan, @phanisrikanth, @prashanttandon, @ps014arkyabagchi, @raihan0001, @raihanrsd, @ratnachand, @rianryu, @ricswhat, @rituljangir, @romeanthem, @ruitongs, @ruwadnaswan, @rwdnaswan, @ryudonghun, @sachinkukkar, @sanjidh090, @saravananj, @sarveshtalele, @seongwonk, @seongwonkang, @shahadsawla, @shakhoyatshujon, @shouta0108, @shyngys19732, @sigmaborov, @simonartzet, @sirteo, @sjskmu, @skssdn, @skuldnorniern, @smak17, @smammahdi, @smammer056, @smcho1201, @snigdhatalasila, @soyaoki, @suam6409, @taejinpark0727, @takamaruex, @tarandeep97, @thiendangnn63, @tjrhyun, @tnr1337, @tonyironman099, @trikang, @varunraskar, @wenjiozhao, @xiaoqiaodong, @xinyvwang, @yashanathaniel, @yeonslay, @yoyy0099, @yuxiazff, @yuxisu944, @zahidbooni
", "votes": null, "replies": [] }, { "id": 3442278, "author_name": "fuumin621", "author_url": "", "post_date": "04/15/2026 08:02:07", "content": "@picekl \nHi, I’m from the winning team of the ACCIDENT challenge. I'm currently preparing a paper for the workshop, but since I'm based in Japan, it's difficult to attend in person. Will there be an option for a virtual or pre-recorded video presentation? Thanks!
", "votes": null, "replies": [ { "id": 3442395, "author_name": "picekl", "author_url": "", "post_date": "04/15/2026 12:37:32", "content": "Hi @fuumin621 ,
\nCongratulations on winning the competition!
\nAbout your question. We are aware of the situation. Thus, we are currently discussing the Workshop format with other Workshop chairs.\nIn any case, virtual participation or prerecorded videos should not be a problem. FYI, I prefer a virtual participation.
\nBest,\nLukas
", "votes": null, "replies": [ { "id": 3442960, "author_name": "fuumin621", "author_url": "", "post_date": "04/16/2026 01:53:46", "content": "Thank you for the clarification! I’m relieved to hear that. I would prefer to participate virtually as well, so I’ll wait for the final details. Thanks again!
", "votes": null, "replies": [] } ] } ] }, { "id": 3444023, "author_name": "yuxiazff", "author_url": "", "post_date": "04/17/2026 06:21:57", "content": "@picekl Hello,I am a member of the team that ranked 2nd in the VQA track. \nI noticed that a \"Call for Papers\" has been posted in the ACCIDENT track discussion, but there hasn't been a similar announcement in the VQA track yet. Therefore, I would like to clarify:\n1.Requirement: Are participants in the VQA track also expected to submit a paper for the workshop?\n2.Format: As we have a very tight schedule currently, would it be acceptable if we submitted a concise technical report instead of a full workshop paper?\n3.Logistics: Could you please provide the submission portal/link and the official deadline for the VQA track?
", "votes": null, "replies": [ { "id": 3445535, "author_name": "picekl", "author_url": "", "post_date": "04/20/2026 02:51:33", "content": "Hi @yuxiazff ,
\nI have no idea. You have to ask the organizers of that competition. \nI have nothing to do with it.
\nGood luck!
\nBest,\nLukas
", "votes": null, "replies": [] } ] }, { "id": 3446314, "author_name": "guangjiandong", "author_url": "", "post_date": "04/21/2026 11:51:05", "content": "Hi, @picekl
\nI noticed that the Molmo-7B results on the leaderboard differ from those reported in Table 5 of the paper (especially for S and ACCS), while the Naive and Best-from-all baselines seem consistent across both.I was wondering if this discrepancy is due to differences in the inference pipeline (e.g., temporal sampling, prompting), dataset versioning, or evaluation protocol between the paper and the leaderboard?
\nWould really appreciate any clarification. Thanks!
", "votes": null, "replies": [ { "id": 3446436, "author_name": "picekl", "author_url": "", "post_date": "04/21/2026 16:02:11", "content": "Hi @guangjiandong ,
\nWhat do you mean by Malmo-7B results on the leaderboard? \nIf you mean This Leaderboard, it is most likely just an old artifact that I'll fix ASAP.
\nIn any case, the results in the paper are 100%; we double-checked them multiple times.
\nBest,\nLukas
", "votes": null, "replies": [] } ] }, { "id": 3449481, "author_name": "varunraskar", "author_url": "", "post_date": "04/28/2026 10:20:16", "content": "@picekl We are trying to make a submission to AUTOPILOT @ CVPR 2026 — Archival Track, but not getting an option to submit it like it's in the Non-Archival Track.
", "votes": null, "replies": [ { "id": 3449524, "author_name": "picekl", "author_url": "", "post_date": "04/28/2026 12:53:47", "content": "Hi @varunraskar ,
\nNot sure what you mean.\nDoes this link not work for you?\nhttps://openreview.net/group?id=thecvf.com/CVPR/2026/Workshop/AUTOPILOT-NA&referrer=%5BHomepage%5D(%2F)#tab-your-consoles
\nBest,\nLukas
", "votes": null, "replies": [ { "id": 3449540, "author_name": "varunraskar", "author_url": "", "post_date": "04/28/2026 13:51:17", "content": "Got it, Thank you
", "votes": null, "replies": [] } ] } ] }, { "id": 3456458, "author_name": "ryudonghun", "author_url": "", "post_date": "05/11/2026 23:39:40", "content": "@picekl Hi, just wanted to follow up on the Non-Archival Track results. The notification was scheduled for May 10th, but I haven't received any updates yet. With the Virtual Materials deadline approaching on May 20th, could you let us know when we might expect the results? Thank you!
", "votes": null, "replies": [ { "id": 3456464, "author_name": "picekl", "author_url": "", "post_date": "05/12/2026 00:09:00", "content": "Hi @ryudonghun ,
\nI just double checked with the other competition organizers and we will send the feedback and notification of acceptance / rejection tomorrow. There are some small issues that caused delay. Sorry for that.
\nBest,\nLukas
", "votes": null, "replies": [ { "id": 3456476, "author_name": "ryudonghun", "author_url": "", "post_date": "05/12/2026 00:55:36", "content": "No worries at all! Thanks for letting me know. Looking forward to it!
", "votes": null, "replies": [] }, { "id": 3457614, "author_name": "ryudonghun", "author_url": "", "post_date": "05/14/2026 02:14:19", "content": "Hi @picekl \nSorry to bother you again with this! I completely understand that the team is busy and things can get delayed. I just wanted to check if there's any update on the notifications, as I really need to start making plans and preparing for my conference participation.\nThank you again for your time and hard work!
", "votes": null, "replies": [ { "id": 3457894, "author_name": "picekl", "author_url": "", "post_date": "05/14/2026 15:12:17", "content": "Hi @ryudonghun ,
\nI have provided inputs to the program chairs on Friday. I think they are waiting for feedback from other competition organizers so they can send everything in one batch.
\nI forwarded your concern, but sadly, there is nothing else I can do at the moment. 😔
\nBest,\nLukas
", "votes": null, "replies": [ { "id": 3457898, "author_name": "ryudonghun", "author_url": "", "post_date": "05/14/2026 15:22:51", "content": "Thank you so much for the transparent update and for advocating on my behalf. I completely understand that it is out of your hands now and that the program chairs need to coordinate everything. I will patiently wait for the official notification. Have a great weekend!
", "votes": null, "replies": [] } ] } ] } ] } ] } ], "raw_markdown_by_id": { "3441401": "As the ACCIDENT challenge has ended, we would like to invite participants to submit their work to the [AUTOPILOT Workshop](https://www.autopilot-cvpr.net/).\n\nTake this as a good opportunity to develop your challenge submission into a workshop paper and present it, as **selected papers will have the opportunity to be presented at CVPR through the AUTOPILOT Workshop.** In addition to reporting final performance, a submission should highlight the main idea behind your approach, design choices, ablations, error analysis, or practical lessons learned during the competition.\n\nWe especially encourage submissions that go beyond leaderboard ranking alone and help others better understand what worked, what did not, and which directions seem promising on the ACCIDENT benchmark.\n\nTo support follow-up work, we have been working hard on several things:\n- Benchmark website with public leaderboards, submission instructions, and a browser-based scoring interface for CSV submissions --> https://accidentbench.github.io/\n- ACCIDENT dataset with labels available on Kaggle --> https://www.kaggle.com/datasets/picekl/accident\n- Dataset paper with full technical details, benchmark design, and evaluation methodology --> https://arxiv.org/abs/2604.09819\n\n**P.S. We are still actively working on the website and related resources. Some parts may still be incomplete or temporarily unavailable as we continue to verify everything and finalize the setup. Please do not use the results returned by the website YET!**", "3441402": "aadigupta1601, @abderrahmen36, @abhaylejith, @adibraian, @adityasiraskar, @ahnaftahmid24, @aira666, @aisang, @aiswarjya, @akumar49, @ameythakur20, @andrewagirgis, @anidiptapal, @anonamename, @apuchakroborti, @arwaabouattia, @asmaasamih22, @asmaasamyh2020, @asmaasamyhagag22, @atomkevich, @barshamishra19, @belkessambadis, @bigshark101, @birunthabanrajendram, @brendayyc, @buharisalisu, @calpis10000, @chancekim, @chayakornwinuntamala, @chengsiqi47, @chenruwen, @choigu, @choisanghyuk, @chow123, @csanskriti, @dhruvpaidukle, @dionysisvoulgarakis, @dipit099, @dkwlsdk, @dla020501, @dqliu12138, @dubbin, @erdespiro, @fariawanpakistan12, @feel12348, @fengsiyi12345678, @fernandollsalvador, @ftakahas, @fuumin621, @gch8678, @gouthamk16, @guangjiandong, @hariprasanth8, @hepingpeace, @hiro356, @hmnshudhmn24, @hongyuchen123321, @hwanghwan, @hwkims, @hyobink, @icebearogo, @ihkali, @irajmoradi, @irisreto, @ishankat, @jayhawk1900, @jek1wantaufik, @jianthuang, @jinchengzhang1, @jooyoungseok, @joshuaburnsblack, @jsalvasoler, @kiyotaka1102, @ksingh51, @kwarkamole, @lakhindarpal, @letita, @lijiaowang97, @lollylandd, @lsiavchangpeng, @lunaandendymion, @mapjoo, @mathurinache, @maximshatskiy, @mdakramhossain1871, @mirzamilanfarabi, @mrpc2003, @ngocnhannguyen, @nhnnguynngc, @novebrititoramadhani, @oliverlang, @parkerke, @patrickiitmz, @patricktz, @peacelu, @pensukesan, @phanisrikanth, @prashanttandon, @ps014arkyabagchi, @raihan0001, @raihanrsd, @ratnachand, @rianryu, @ricswhat, @rituljangir, @romeanthem, @ruitongs, @ruwadnaswan, @rwdnaswan, @ryudonghun, @sachinkukkar, @sanjidh090, @saravananj, @sarveshtalele, @seongwonk, @seongwonkang, @shahadsawla, @shakhoyatshujon, @shouta0108, @shyngys19732, @sigmaborov, @simonartzet, @sirteo, @sjskmu, @skssdn, @skuldnorniern, @smak17, @smammahdi, @smammer056, @smcho1201, @snigdhatalasila, @soyaoki, @suam6409, @taejinpark0727, @takamaruex, @tarandeep97, @thiendangnn63, @tjrhyun, @tnr1337, @tonyironman099, @trikang, @varunraskar, @wenjiozhao, @xiaoqiaodong, @xinyvwang, @yashanathaniel, @yeonslay, @yoyy0099, @yuxiazff, @yuxisu944, @zahidbooni", "3442278": "picekl \nHi, I’m from the winning team of the ACCIDENT challenge. I'm currently preparing a paper for the workshop, but since I'm based in Japan, it's difficult to attend in person. Will there be an option for a virtual or pre-recorded video presentation? Thanks!", "3442395": "Hi @fuumin621 ,\n\nCongratulations on winning the competition!\n\nAbout your question. We are aware of the situation. Thus, we are currently discussing the Workshop format with other Workshop chairs.\nIn any case, virtual participation or prerecorded videos should not be a problem. FYI, I prefer a virtual participation.\n\nBest,\nLukas", "3442960": "Thank you for the clarification! I’m relieved to hear that. I would prefer to participate virtually as well, so I’ll wait for the final details. Thanks again!", "3444023": "picekl Hello,I am a member of the team that ranked 2nd in the VQA track. \nI noticed that a \"Call for Papers\" has been posted in the ACCIDENT track discussion, but there hasn't been a similar announcement in the VQA track yet. Therefore, I would like to clarify:\n1.Requirement: Are participants in the VQA track also expected to submit a paper for the workshop?\n2.Format: As we have a very tight schedule currently, would it be acceptable if we submitted a concise technical report instead of a full workshop paper?\n3.Logistics: Could you please provide the submission portal/link and the official deadline for the VQA track?", "3445535": "Hi @yuxiazff ,\n\nI have no idea. You have to ask the organizers of that competition. \nI have nothing to do with it. \n\nGood luck!\n\nBest,\nLukas", "3446314": "Hi, @picekl \n\nI noticed that the Molmo-7B results on the leaderboard differ from those reported in Table 5 of the paper (especially for S and ACCS), while the Naive and Best-from-all baselines seem consistent across both.I was wondering if this discrepancy is due to differences in the inference pipeline (e.g., temporal sampling, prompting), dataset versioning, or evaluation protocol between the paper and the leaderboard?\n\nWould really appreciate any clarification. Thanks!", "3446436": "Hi @guangjiandong ,\n\nWhat do you mean by Malmo-7B results on the leaderboard? \nIf you mean [This Leaderboard](https://accidentbench.github.io/benchmarks/zero-shot/index.html), it is most likely just an old artifact that I'll fix ASAP.\n\nIn any case, the results in the paper are 100%; we double-checked them multiple times.\n\nBest,\nLukas", "3449481": "picekl We are trying to make a submission to AUTOPILOT @ CVPR 2026 — Archival Track, but not getting an option to submit it like it's in the Non-Archival Track.", "3449524": "Hi @varunraskar ,\n\nNot sure what you mean.\nDoes this link not work for you?\nhttps://openreview.net/group?id=thecvf.com/CVPR/2026/Workshop/AUTOPILOT-NA&referrer=%5BHomepage%5D(%2F)#tab-your-consoles\n\nBest,\nLukas", "3449540": "Got it, Thank you", "3456458": "picekl Hi, just wanted to follow up on the Non-Archival Track results. The notification was scheduled for May 10th, but I haven't received any updates yet. With the Virtual Materials deadline approaching on May 20th, could you let us know when we might expect the results? Thank you!", "3456464": "Hi @ryudonghun ,\n\nI just double checked with the other competition organizers and we will send the feedback and notification of acceptance / rejection tomorrow. There are some small issues that caused delay. Sorry for that.\n\nBest,\nLukas", "3456476": "No worries at all! Thanks for letting me know. Looking forward to it!", "3457614": "Hi @picekl \nSorry to bother you again with this! I completely understand that the team is busy and things can get delayed. I just wanted to check if there's any update on the notifications, as I really need to start making plans and preparing for my conference participation.\nThank you again for your time and hard work!", "3457894": "Hi @ryudonghun ,\n\nI have provided inputs to the program chairs on Friday. I think they are waiting for feedback from other competition organizers so they can send everything in one batch. \n\nI forwarded your concern, but sadly, there is nothing else I can do at the moment. 😔\n\nBest,\nLukas", "3457898": "Thank you so much for the transparent update and for advocating on my behalf. I completely understand that it is out of your hands now and that the program chairs need to coordinate everything. I will patiently wait for the official notification. Have a great weekend!" }, "source": "meta" }