Jaringan Syaraf Tiruan Berbasis Wilayah: Perbedaan antara revisi

* November 2013: '''R-CNN'''. GivenDiberikan ansebuah inputgambar imageinput, R-CNN beginsdimulai bydengan applyingmenerapkan amekanisme mechanismyang calleddisebut SelectivePencarian SearchSelektif tountuk extractmengekstrak [[Region of interest|regions of interest]] (ROI), wheredi eachmana setiap ROI isadalah asebuah rectanglepersegi thatpanjang mayyang representdapat themerepresentasikan boundarybatas ofsebuah anobjek objectdalam in imagegambar. Depending onTergantung thepada scenarioskenarionya, theremungkin mayada besebanyak asdua manyribu as two thousand ROIsROI. AfterSetelah thatitu, eachsetiap ROI isdimasukkan fedmelalui throughjaringan asyaraf neuraluntuk networkmenghasilkan tofitur producekeluaran. outputUntuk features.setiap Forfitur eachkeluaran ROI's output features, akumpulan collection ofpengklasifikasi [[support-vectormesin vektor machinependukung]] classifiersdigunakan isuntuk usedmenentukan tojenis determine what type of objectobjek (ifjika anyada) is containedyang withinterkandung thedalam ROI.<ref>{{Cite news|last=Gandhi|first=Rohith|url=https://towardsdatascience.com/r-cnn-fast-r-cnn-faster-r-cnn-yolo-object-detection-algorithms-36d53571365e|title=R-CNN, Fast R-CNN, Faster R-CNN, YOLO — Object Detection Algorithms|date=July9 9,Juli 2018|work=Towards Data Science|access-date=March12 12,Maret 2020}}</ref>

* April 2015: '''Fast R-CNN'''. While the originalSementara R-CNN independentlyyang computedasli thesecara neuralindependen networkmenghitung featuresfitur onjaringan eachsaraf ofpada asmasing-masing manysebanyak asdua tworibu thousandwilayah regionsyang of interestdiminati, Fast R-CNN runsmenjalankan thejaringan neuralsaraf networksatu oncekali onpada theseluruh whole imagegambar. AtPada theakhir endjaringan ofterdapat themetode networkbaru isyang a novel method calleddisebut ROIPooling, whichyang slicesmemotong out eachsetiap ROI fromdari thetensor network'skeluaran output tensorjaringan, reshapesmembentuk itulang, anddan classifies itmengklasifikasikannya. AsSeperti in the originalpada R-CNN asli, the Fast R-CNN usesmenggunakan SelectivePencarian SearchSelektif tountuk generatemenghasilkan itsproposal region proposalswilayahnya.<ref name=":0">{{Cite news|last=Bhatia|first=Richa|url=https://analyticsindiamag.com/what-is-region-of-interest-pooling/|title=What is region of interest pooling?|date=10 September 10, 2018|work=Analytics India|access-date=March12 12,Maret 2020}}</ref>

* JuneJuni 2015: '''Faster R-CNN'''. WhileSementara Fast R-CNN usedmenggunakan SelectivePencarian SearchSelektif tountuk generatemenghasilkan ROIsROI, Faster R-CNN integratesmengintegrasikan thegenerasi ROI generationke intodalam thejaringan neuralsaraf networkitu itselfsendiri.<ref name=":0" />

* MarchMaret 2017: '''Mask R-CNN'''. WhileSementara previous versions ofversi R-CNN focusedsebelumnya onberfokus objectpada deteksi detectionobjek, Mask R-CNN addsmenambahkan instancesegmentasi segmentationinstance. Mask R-CNN alsojuga replacedmenggantikan ROIPooling withdengan ametode newbaru methodyang calleddisebut ROIAlign, whichyang candapat representmerepresentasikan fractionspecahan of a pixelpiksel.<ref>{{Cite news|last=Farooq|first=Umer|url=https://medium.com/@umerfarooq_26378/from-r-cnn-to-mask-r-cnn-d6367b196cfd|title=From R-CNN to Mask R-CNN|date=February 15, 2018|work=Medium|access-date=March12 12,Maret 2020}}</ref><ref>{{Cite news|last=Weng|first=Lilian|url=https://lilianweng.github.io/lil-log/2017/12/31/object-recognition-for-dummies-part-3.html|title=Object Detection for Dummies Part 3: R-CNN Family|date=December31 31,Desember 2017|work=Lil'Log|access-date=March12 12,Maret 2020}}</ref>

* JuneJuni 2019: '''Mesh R-CNN''' addsmenambahkan thekemampuan abilityuntuk tomenghasilkan generate amesh 3D meshdari from agambar 2D image.<ref>{{Cite news|last=Wiggers|first=Kyle|url=https://venturebeat.com/2019/10/29/facebook-highlights-ai-that-converts-2d-objects-into-3d-shapes/|title=Facebook highlights AI that converts 2D objects into 3D shapes|date=October 29, 2019|work=VentureBeat|access-date=March 12, 2020}}</ref>

Region-basedJaringan convolutionalsyaraf neuraltiruan networksberbasis havewilayah beentelah useddigunakan foruntuk trackingmelacak objectsobjek fromdari akamera drone-mountedyang cameradipasang di [[pesawat nirawak]],<ref>{{Cite news|last=Nene|first=Vidi|url=https://dronebelow.com/2019/08/02/deep-learning-based-real-time-multiple-object-detection-and-tracking-via-drone/|title=Deep Learning-Based Real-Time Multiple-Object Detection and Tracking via Drone|date=Aug2 2,Agustus 2019|work=Drone Below|access-date=Mar28 28,Maret 2020}}</ref> locating text in an image,<ref>{{Cite news|last=Ray|first=Tiernan|url=https://www.zdnet.com/article/facebook-pumps-up-character-recognition-to-mine-memes/|title=Facebook pumps up character recognition to mine memes|date=Sep 11, 2018 |publisher=[[ZDNET]] |access-date=Mar 28, 2020}}</ref> anddan enablingmemungkinkan objectpendeteksian detectionobjek indi [[Google Lens]].<ref>{{Cite news|last=Sagar|first=Ram|url=https://analyticsindiamag.com/these-machine-learning-techniques-make-google-lens-a-success/|title=These machine learning methods make google lens a success|date=Sep 9, 2019|work=Analytics India|access-date=Mar 28, 2020}}</ref> Mask R-CNN servesberfungsi assebagai onesalah ofsatu sevendari taskstujuh intugas thedalam MLPerf Training Benchmark, whichyang ismerupakan a competition to speed upkompetisi theuntuk trainingmempercepat ofpelatihan neuraljaringan networkssaraf.<ref>{{cite arXiv|eprint=1910.01500v3|class=math.LG|first=Peter|last=Mattson|title=MLPerf Training Benchmark|date=2019|display-authors=etal}}</ref>