作者查詢 / followwar
作者 followwar 在 PTT [ DataScience ] 看板的留言(推文), 共44則
限定看板：DataScience
看板排序：
全部Gossiping7859PuzzleDragon2693Stock1295CFantasy812Beauty784nobunyaga464model435Tech_Job337LoL231PathofExile212joke201SportLottery165sex140WorldCup110Boy-Girl109RO103CSMU-Psy89movie78Soft_Job75Cheer64ToS64Lottery63Hearthstone52specialman47DataScience44AnimalForest40AfterPhD32C_Chat27Militarylife25NSwitch24Mabinogi23PhD22HatePolitics18KS98-30218StupidClown18WomenTalk18Salary16UmaMusume16Baseball14DNF14FuMouDiscuss12PublicIssue12FLAT_CLUB11HCHS_ALUMNI11L_TalkandCha11marvel11MH11TigerBlue11PC_Shopping10Physics10AC_In9Golden-Award9graduate9kodomo9NTU9Tainan9Teacher9AllTogether8cat8Hunter8NCHU-CE-SB8paranormal8Doraemon7Emergency7HCHS583187HCHS921047HCHS923077HCSHch13_3117YOLO7GUNNM6HK-movie6Jam6KOTDFansClub6LTK6MenTalk6PCSH91_3056PttHistory6RFonline6Aves5Dragonica5Jolin5Lineage5NSYSU_MIS_995NTUBIME-1025DragonBall4FEZ4home-sale4HSNU_8204I-Lan4Kaohsiung4KS97-3164Miaoli4NKUTEE4Olympics_ISG4PlayStation4Sheena4SouthPark4THU_BA20004TuCheng4Wanhua4YUYUHAKUSHO4ASIA-uni3car3CCRomance3Cognitive3DYU3FCU_Talk3Football3GatoShoji3GUNDAM3HCHS933143HSNU_11153Hualien3iOS3Keelung3KS95-3013KS95-3183KS97-3103KS_PMAC3LordsOfWater3marriage3medstudent3NTUCivilism3NTUmed003NUK_EE100A3NUU_Talk3Paradox3ScienceNote3SENIORHIGH3SMSlife3Sub_GMobile3TKU_EE_92C3TuTsau3WorkinChina3Zombie3ArakawaCow2BigSanchung2biker2BioHazard2Bioindustry2book2Buddhism2CareerPlan2ChainChron2ckbc2cksh85th3172CLHS-50-142CSMU-HSA952Expansion072fastfood2FCUMCAE-SB2FengYuan2Finance2gay2GemTang2GossipPicket2GuessX32HCHS593052Hiking2Horror2Hsinchu2IA2ID_Multi2Japan_Travel2japanavgirls2JinYong2junji-ITO2kawaii2KERORO2Kids_Sucker2Kindaichi_Q2KingdomHuang2KS93-3152KS95-3022KS95-3162KS97-3202LSSH2Math2medache2MOD2MONSTER2MuscleBeach2NBA2NCKU_EARTH982NCKU_PT962NCTU_TALK2NCU97ME-B2NCYU_Fst_982NSYSU2NTOU_TALK2NTUST-ENG2NTUST_Talk2Old-Games2PokeMon2Post2PushDoll2rent-exp2RIPE_gender2SCU_CIS-92A2Shu-Lin2ShuangHe2Steam2SuperHeroes2SYSOP2TaichungBun2Taitung2TAKESHI2talk2TamShui2THULAW2TKU_IHM2TW-history2UEFA2Weyslii2Wine2YiGo3112ADS1AKB481Anchors1AngelPray1ANIMAX1AnimMovie1AOCSCV93B1AOE1ArabicSoccer1ASHS-93-li1ASHS-95RN1AT_PingPong1Atheism1b964060XX1B97A013XX1B_supporters1balaz1Balloon1battlegirlhs1BBSview1BCC_Midnight1BigPeitou1BigShiLin1Biotech1Bread1Browsers1CATCH1Catholic1CCHS-98011CCSH_86_3121CCU-CHEMICAL1CD-R1CGSH86th3231CGSH88th3081Chan_Mou1ChangHua1Childhood1Christianity1CHSH-3191ck53rd3281ck56th3221CKBIGBOAT1cksh84th3021CKSHTalk1Coffee1CSMU-HSA961CSMU-MED971CSMU-MIS961CSMU-MT991CTSH963011CTSH97EXP1CVS1CYCU_ICE90AB1DeathNote1DIABLO1DiscoveryNGC1dlsh-7th-3031DPP1DragonQuest1Drink1DummyHistory1EAseries1EatToDie1Ecophilia1Education1ESP1eSports1F_Naohito1fashion1FCU_EE97A1feminine_sex1Feminism1FinalFantasy1FiremanLife1fivesix1FJU1FJU-ECE-20071FJU-EE-2003b1FJU-EE-VLSI1FJU-Laws971FJU-Stat95A1FJU_Chiayun1FJU_Fashion1FongShan1fruitbasket1FSHS-96-3121FujisakiRyu1gallantry1Geography1Gintama1GO1GoodPregnan1GreenParty1GVOnline1Haruhi1Hayate1Heroes1Hip-Hop1historia1HomeTeach1HSNU_10081HSNU_10631HSNU_11121HSNU_11421HSNU_11461HSNU_9751HSNU_J1231ILSH-963131Inference1Instant_Food1Isayama1JapanStudy1Jay1JP_Entertain1JPliterature1KanColle1kartrider1Keanu1KIDs1Kings1Kitty_Sanrio1KMT1Knicks1KoreaDrama1KoreaStar1KS90-3031KS90-3091KS91-3051KS91-3191KS92-3121KS94-3171KS94-3201KS94-3211KS95-2061KS95-3061KS95-3071KS95-3101KS96-2021KS96-3011KS96-3051KS96-3141KS97-2161KS97-3011KS97-3021KS97-3081KS97-3141KS97-3181KS98-3141KSHS_Talk1KUAS_ME94A1Law-Fala1LawsuitSug1Lifeismoney1LightNovel1LittleNew1LivingGoods1LUCHUO3321Magic1MartialArts1MCJH_M_081MeatFamily1MiamiHeat1Military1MLBGAME1mobile-game1MobileComm1money1MP1Mudran1MysticWiz1NCCU05_GIDS1NCCU08_Ethno1NCCU08_SW1NCCU09_LawA1NCCU09_TWHis1NCCU_C-Baske1NCCU_SEED1NCCU_TEA1NCHU-FS981NCHU-MMA1NCHU-photo1NCHU_AMpower1NCHUS1NCKU-physiol1NCU91Finance1NCUECON961NDHU-His1021NDMC-PH271Neihu1NightLife1NKNU_Phy1001North_Coast1NSYSU_MSOE011NSYSU_MSOE991NTHU_IEEM-941NTHU_pingpon1NTNU-HisSB1NTPU-ACC-BA1NTPU-CSIE941NTTU_CSIE991NTU-Archery1NTU-NeiHu1NTUE-CS1021NTUE-CS981NTUEE1111NTUEE1151NTUGIEE_AMTG1NTUGIPO_PNSL1NTUmed011NTUMiaoli1NTUST-EE-B951NTUST-TX-B951NTUSTfinM971Nubee1NUK-APIBM1001NUK_AC981NUK_TALK1NUTN_MS991NUU_EE1NY-Yankees1NYUST1ONE_PIECE1Ornithology1Orzhong97cl1part-time1PCCU-CS1PCSH92_3091PCSH96_3061PHAT-MEN1PingTung1Pisces1PokemonGO1politics1Printer3D1Ptt-Charity1PttLifeLaw1PublicServan1Railway1RealDream1Reptile1RMI_Soccer1Road1Road_Running1S.H.E1SAN-YanYi1San-Ying1SCU_ACC_96A1SilentHill1SKET_DANCE1sky1SongShan1SRW1SSSH-11th3041SSSH-16th3131StarWars1Statistics1Stephen1StockPicket1studyabroad1studyteacher1Sub_GOnline1Sub_RolePlay1SW_Job1TaiShan1Taoyuan1Tennis1tennis_life1TeresaTeng1textbook1TFSHS68th3121theshine41THU-CHE961TKU_BF93C1TKU_EW94B1TNFSH102nd1Tobacco1toberich1Transport981TunHua05t3111twin1TY_Research1Tyukaitiban1UTAH-JAZZ1Viator96Chia1Viator96Gi1Vocaloid1W-Philosophy1WarCraft1wearefriends1WuLing50-3131XBOX1XiangSheng1YAKYU1YoungDotx31YP94-3041YZU_CN99A1Zastrology1<< 收起看板(513)
3F→: 裝CUDA10自己bulid pytorch08/17 20:47
29F→: 先說是那些研究工作這樣做?08/15 22:09
10F→: batch normalization在training時只用自己的mean std07/20 14:34
11F→: 哪裡來的超出一個batch的資訊...07/20 14:34
12F→: *精準一點 只用自己目前batch的mean std07/20 14:35
13F→: 是 intra-batch information leakage07/20 14:37
14F→: 因為training時mean std是整個batch算07/20 14:39
15F→: 可以藉由這個資訊學到sample之間的不同07/20 14:39
16F→: 由於contrastive learning是要學自己跟別人不一樣07/20 14:40
17F→: BN偷偷幫助了這件事07/20 14:41
18F→: 所以變成model藉由BN來區分 本身並沒有習得資訊07/20 14:42
9F→: 先想想你的CAM作法真的能看出CNN做了什麼嗎? (笑)05/18 22:59
10F→: 不要太一廂情願的覺得CAM一定對05/18 23:00
13F→: 你可以參考一下 Grad-CAM Score-CAM 之類的文章05/21 14:22
20F→: 你對資料取的features決定分群的好壞02/22 22:17
21F→: features反映什麼特性 就分出啥02/22 22:18
2F→: 看資料難易度 VOC沒有0.8就慘 COCO有0.4還蠻屌的11/07 20:59
14F→: pytorch makes life easier XD10/25 21:40
6F→: 看看一些 fine-grained的paper吧....10/12 23:31
1F→: styleGAN V206/04 09:47
8F→: trick全上 幫你試一遍坑 算良心之作了吧05/01 02:03