Thank you Foivos. Please email a link to your downloadable recording to @[email protected]<mailto:[email protected]>.
From: Foivos Diakogiannis <[email protected]> Date: Monday, November 30, 2020 at 11:26 PM To: Vartika Singh <[email protected]>, Foivos Diakogiannis <[email protected]> Cc: Triston Cao <[email protected]>, Sam Skalicky <[email protected]>, "Zha, Sheng" <[email protected]> Subject: Re: Present at Apache MXNet Day 2020 on Looking for change? Roll the Dice and demand Attention External email: Use caution opening links or attachments Dear Vartika and all, thank you, this is a great honour. I will send you video of the talk by the requested date. Kind regards, Foivos ---------------------------------------------------------------------- Dr. Foivos I. Diakogiannis Senior Research Fellow - Data Science ICRAR - The University of Western Australia P: +61 8 6488 7206 [cid:[email protected]] https://twitter.com/Foivos_Diak [cid:[email protected]] https://www.linkedin.com/in/foivos-diakogiannis/ [cid:[email protected]] https://github.com/feevos/ [cid:[email protected]] www.icrar.org<http://www.icrar.org/> ________________________________ From: Vartika Singh <[email protected]> Sent: 01 December 2020 14:19 To: Foivos Diakogiannis <[email protected]>; Foivos Diakogiannis <[email protected]> Cc: Triston Cao <[email protected]>; Sam Skalicky <[email protected]>; Zha, Sheng <[email protected]> Subject: Present at Apache MXNet Day 2020 on Looking for change? Roll the Dice and demand Attention Hello Foivos, The Apache MXNet team has accepted your talk. We are very excited to have you present on this event. Your talk has been scheduled at 10:31 AM PST on Dec 14th for 15 minutes. The agenda/time is still evolving. We will inform you if there are any changes. Here is the updated website - https://s.apache.org/MXNet2020<https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fs.apache.org%2FMXNet2020&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817825788%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=WZb73%2B%2BUjkFxkFXtzIe9OhVtZvbm5O6U2rirQEohmjI%3D&reserved=0> Action Required: * We request you to send your recorded talk (15 minutes) to us by Dec 6th midnight PST. * Please respond by end of tomorrow to confirm your presentation and attendance. Question Answers: We plan to have a slack channel in parallel to your recording where you can respond to questions while your recording plays. We will add the slack channel names to the respective talks in a week. Please reach out if you have any questions. We will send out the webcast details by end of Wednesday. Thank you Vartika From: Foivos Diakogiannis <[email protected]> Date: Tuesday, November 24, 2020 at 6:32 PM To: apachemxnetday <[email protected]>, Foivos Diakogiannis <[email protected]> Subject: Re: Application for presentation Dear Vertika, thank you for your email. Please find below the information you requested. 1. Short Bio: Foivos I. Diakogiannis graduated from the University of Athens in 2004. He completed his Ph.D. in 2014 (University of Sydney) on (computational) Astronomy. His work was on dark matter mass estimates of dwarf spheroidal galaxies. In 2014 he moved to WA for his first PostDoc on ICRAR/UWA (Astronomy). In 2017 he changed fields, from Astronomy to deep learning and completed a 2nd postdoc in Data61/CSIRO, focusing on deep learning applications to Remote Sensing problems (computer vision). He is currently working as a senior research fellow (data science) in the International Centre for Radio Astronomy Research (the University of Western Australia). Social media links: https://www.linkedin.com/in/foivos-diakogiannis/<https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.linkedin.com%2Fin%2Ffoivos-diakogiannis%2F&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817825788%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=7eBwR5QmO1E2YYDO8vW1KxrCnM0eJMyzufSUrbfmIZo%3D&reserved=0> <https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Ftwitter.com%2FFoivos_Diak&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817835776%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=HW2iDkrVytsC0uz9In0wr%2FW72Eswg6kGjYJ61DHsKng%3D&reserved=0> https://twitter.com/Foivos_Diak<https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Ftwitter.com%2FFoivos_Diak&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817835776%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=HW2iDkrVytsC0uz9In0wr%2FW72Eswg6kGjYJ61DHsKng%3D&reserved=0> https://github.com/feevos<https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Ffeevos&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817845771%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=Ypt2hEG3zlLcP5%2F5gV%2BlMAS%2BpziRIxm4Ai56jAPweWc%3D&reserved=0> 1. Can do either 15min or 30min, it really depends on the preference of the organisers. If it is 15min I will only talk about change detection, if it is 30min I will also enrich with semantic segmentation from our past work. I have given a similar talk (30+min) in the murdoch university quite recently, and you can find the slides of the talk here: https://github.com/feevos/ceecnet/blob/master/doc/csse_talk_diakogiannis.pdf<https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Ffeevos%2Fceecnet%2Fblob%2Fmaster%2Fdoc%2Fcsse_talk_diakogiannis.pdf&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817845771%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=JcfbdHNWB%2BcN5%2FDxPmsryUtDi3MbcvnyH1oeCpgZcqk%3D&reserved=0> Kind regards, Foivos ---------------------------------------------------------------------- Dr. Foivos I. Diakogiannis Senior Research Fellow - Data Science ICRAR - The University of Western Australia P: +61 8 6488 7206 [cid:[email protected]] https://twitter.com/Foivos_Diak<https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Ftwitter.com%2FFoivos_Diak&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817855764%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=k9thCCuqSFlIgkaGI6fmr%2FRUeiRqq81xZl9vrx1MTqc%3D&reserved=0> [cid:[email protected]] https://www.linkedin.com/in/foivos-diakogiannis/<https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.linkedin.com%2Fin%2Ffoivos-diakogiannis%2F&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817855764%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=p3j41puEnI8CY8LotYhIKWf0DM40IK74VIOxtrF2ZQA%3D&reserved=0> [cid:[email protected]] https://github.com/feevos/<https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Ffeevos%2F&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817865762%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=a%2BsoBUSbRWBfbEHFsRTOXU4Ni4N9iCiMjQ1FRZyDn%2FQ%3D&reserved=0> [cid:[email protected]] www.icrar.org<https://aus01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.icrar.org%2F&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817865762%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=LvHtnEAdhpAZht48VCYbde0YTTpTpyJCQbO31MOVnjU%3D&reserved=0> ________________________________ From: apachemxnetday <[email protected]> Sent: 25 November 2020 10:10 To: Foivos Diakogiannis <[email protected]> Cc: Foivos Diakogiannis <[email protected]> Subject: Re: Application for presentation Hello Foivos, Thank you for submitting the talk to Apache MXNet Day. I have a few questions for you if you could respond before tomorrow, it will be very helpful. 1. Brief user bio/profile with relevant social links (LinkedIn, Twitter what have you) 2. Would you prefer a 30 minute slot or a 15 minute one. Our review committee will make a decision on this, but we think it will be helpful to have your guidance as well. Thank you again Vartika From: Foivos Diakogiannis <[email protected]> Date: Saturday, November 14, 2020 at 9:32 PM To: apachemxnetday <[email protected]> Cc: Foivos Diakogiannis <[email protected]> Subject: Application for presentation Dear all, I would please like to be considered for presenting on the apache mxnet day, in the category "Research and applications". I assume that the mxnet day will also include applications written with apache mxnet that deviate from computer science - hence my application. If yes, then you may find the topic of my presentation interesting: I will present our latest work on change detection ( https://arxiv.org/abs/2009.02062<https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Farxiv.org%2Fabs%2F2009.02062&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817875750%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=GUJoi9x3aR0lYd5aBFZFZ9dymWpxfUKMBN09f%2Bl%2FsTE%3D&reserved=0> - under review, https://github.com/feevos/ceecnet<https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Ffeevos%2Fceecnet&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817875750%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=xtV48Hn%2FjGr1wPKgfFAbIC51Bgt42Bnh2yUl4hkNMiU%3D&reserved=0> ). The talk will be focused on change detection, with some highlights from our previous research on semantic segmentation (resunet-a https://www.sciencedirect.com/science/article/pii/S0924271620300149<https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS0924271620300149&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817885745%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=nJw9%2FnNW1QwvnMCa89SjO9ZjnnT2dM3vwqLK%2FLhYh98%3D&reserved=0> , https://github.com/feevos/resuneta<https://aus01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Ffeevos%2Fresuneta&data=04%7C01%7Cfoivos.diakogiannis%40uwa.edu.au%7Cb72461359ff64567c4fc08d895c1159b%7C05894af0cb2846d8871674cdb46e2226%7C1%7C1%7C637424003817885745%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=s7c3mqjuTssQtgQdFzMt1pZ2nmXQgWBUrHgAyTWKou0%3D&reserved=0>), all built with Apache mxnet. Title: Looking for change? Roll the Dice and demand Attention Abstract: Change detection, i.e. identification per pixel of changes for some classes of interest from a set of bi-temporal co-registered images,is a fundamental task in the field of remote sensing. It remains challenging due to unrelated forms of change that appear at different times in input images. These are changes due to different environmental conditions or simply changes of objects that are not of interest. Here, we propose a reliable deep learning framework for the task of semantic change detection in very high-resolution aerial images. Our framework consists of a new loss function, new attention modules, new feature extraction building blocks, anda new backbone architecture that is tailored for the task of semantic change detection. Specifically, we define a new form of set similarity, that is based on an iterative evaluation of a variant of the Dice coefficient. We use this similarity metric to define a new loss function as well as a new spatial and channel convolution Attention layer (the FracTAL). The new attention layer, designed specifically for vision tasks, is memory efficient, thus suitable for use in all levels of deep convolutional networks. Based on these,we introduce two new efficient self-contained feature extraction convolution units. We term these units CEECNet and FracTAL ResNet units. We validate the performance of these feature extraction building blocks on the CIFAR10 reference data and compare the results with standard ResNet modules. Further, we introduce a new encoder/decoder scheme, a network macro-topology, that is tailored for the task of change detection. We validate our approach by showing excellent performance and achieving state of the art score (F1 and Intersection over Union - hereafter IoU) on two building change detection datasets, namely, the LEVIRCD (F1:0.918, IoU: 0.848) and the WHU (F1: 0.938, IoU: 0.882) datasets Kind regards, Foivos I. Diakogiannis ---------------------------------------------------------------------------- Dr. Foivos I. Diakogiannis Senior Research Fellow - Data Science ICRAR - The University of Western Australia P: +61 8 6488 7206 E: [email protected]<mailto:[email protected]>
