FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2893822766> ?p ?o ?g. }

• W2893822766 abstract "We propose Deeply Supervised Object Detectors (DSOD), an object detection framework that can be trained from scratch. Recent advances in object detection heavily depend on the off-the-shelf models pre-trained on large-scale classification datasets like ImageNet and OpenImage. However, one problem is that adopting pre-trained models from classification to detection task may incur learning bias due to the different objective function and diverse distributions of object categories. Techniques like fine-tuning on detection task could alleviate this issue to some extent but are still not fundamental. Furthermore, transferring these pre-trained models across discrepant domains will be more difficult (e.g., from RGB to depth images). Thus, a better solution to handle these critical problems is to train object detectors from scratch, which motivates our proposed method. Previous efforts on this direction mainly failed by reasons of the limited training data and naive backbone network structures for object detection. In DSOD, we contribute a set of design principles for learning object detectors from scratch. One of the key principles is the deep supervision, enabled by layer-wise dense connections in both backbone networks and prediction layers, plays a critical role in learning good detectors from scratch. After involving several other principles, we build our DSOD based on the single-shot detection framework (SSD). We evaluate our method on PASCAL VOC 2007, 2012 and COCO datasets. DSOD achieves consistently better results than the state-of-the-art methods with much more compact models. Specifically, DSOD outperforms baseline method SSD on all three benchmarks, while requiring only 1/2 parameters. We also observe that DSOD can achieve comparable/slightly better results than Mask RCNN + FPN (under similar input size) with only 1/3 parameters, using no extra data or pre-trained models." @default.
• W2893822766 created "2018-10-05" @default.
• W2893822766 creator A5003418019 @default.
• W2893822766 creator A5026310549 @default.
• W2893822766 creator A5031514811 @default.
• W2893822766 creator A5047962986 @default.
• W2893822766 creator A5049353316 @default.
• W2893822766 creator A5084452813 @default.
• W2893822766 date "2018-09-24" @default.
• W2893822766 modified "2023-10-17" @default.
• W2893822766 title "Object Detection from Scratch with Deep Supervision" @default.
• W2893822766 cites W1514535095 @default.
• W2893822766 cites W1533861849 @default.
• W2893822766 cites W1536680647 @default.
• W2893822766 cites W1817277359 @default.
• W2893822766 cites W1832500336 @default.
• W2893822766 cites W1861492603 @default.
• W2893822766 cites W1898560071 @default.
• W2893822766 cites W1903029394 @default.
• W2893822766 cites W1931639407 @default.
• W2893822766 cites W1947481528 @default.
• W2893822766 cites W1948751323 @default.
• W2893822766 cites W2031489346 @default.
• W2893822766 cites W2088049833 @default.
• W2893822766 cites W2095705004 @default.
• W2893822766 cites W2097117768 @default.
• W2893822766 cites W2102605133 @default.
• W2893822766 cites W2104657103 @default.
• W2893822766 cites W2108598243 @default.
• W2893822766 cites W2140609507 @default.
• W2893822766 cites W2155893237 @default.
• W2893822766 cites W2161381512 @default.
• W2893822766 cites W2163605009 @default.
• W2893822766 cites W2183341477 @default.
• W2893822766 cites W2194775991 @default.
• W2893822766 cites W2207849498 @default.
• W2893822766 cites W2274287116 @default.
• W2893822766 cites W2288122362 @default.
• W2893822766 cites W2302255633 @default.
• W2893822766 cites W2407521645 @default.
• W2893822766 cites W2515655118 @default.
• W2893822766 cites W2557728737 @default.
• W2893822766 cites W2565639579 @default.
• W2893822766 cites W2570343428 @default.
• W2893822766 cites W2607119937 @default.
• W2893822766 cites W2613718673 @default.
• W2893822766 cites W2768489488 @default.
• W2893822766 cites W2772989637 @default.
• W2893822766 cites W2796347433 @default.
• W2893822766 cites W2798355657 @default.
• W2893822766 cites W2811135961 @default.
• W2893822766 cites W2883208628 @default.
• W2893822766 cites W2894400606 @default.
• W2893822766 cites W2901394229 @default.
• W2893822766 cites W2913414428 @default.
• W2893822766 cites W2949117887 @default.
• W2893822766 cites W2950178297 @default.
• W2893822766 cites W2950975557 @default.
• W2893822766 cites W2962835968 @default.
• W2893822766 cites W2962851801 @default.
• W2893822766 cites W2962917547 @default.
• W2893822766 cites W2962992847 @default.
• W2893822766 cites W2963016543 @default.
• W2893822766 cites W2963037989 @default.
• W2893822766 cites W2963150001 @default.
• W2893822766 cites W2963150697 @default.
• W2893822766 cites W2963342610 @default.
• W2893822766 cites W2963351448 @default.
• W2893822766 cites W2963446712 @default.
• W2893822766 cites W2963604034 @default.
• W2893822766 cites W2963758027 @default.
• W2893822766 cites W2963786238 @default.
• W2893822766 cites W2963813458 @default.
• W2893822766 cites W2963840672 @default.
• W2893822766 cites W2963927307 @default.
• W2893822766 cites W2964080601 @default.
• W2893822766 cites W2964241181 @default.
• W2893822766 cites W2964288706 @default.
• W2893822766 cites W2964299589 @default.
• W2893822766 cites W3012573144 @default.
• W2893822766 cites W3106250896 @default.
• W2893822766 cites W753847829 @default.
• W2893822766 cites W845365781 @default.
• W2893822766 doi "https://doi.org/10.48550/arxiv.1809.09294" @default.
• W2893822766 hasPublicationYear "2018" @default.
• W2893822766 type Work @default.
• W2893822766 sameAs 2893822766 @default.
• W2893822766 citedByCount "0" @default.
• W2893822766 crossrefType "posted-content" @default.
• W2893822766 hasAuthorship W2893822766A5003418019 @default.
• W2893822766 hasAuthorship W2893822766A5026310549 @default.
• W2893822766 hasAuthorship W2893822766A5031514811 @default.
• W2893822766 hasAuthorship W2893822766A5047962986 @default.
• W2893822766 hasAuthorship W2893822766A5049353316 @default.
• W2893822766 hasAuthorship W2893822766A5084452813 @default.
• W2893822766 hasBestOaLocation W28938227661 @default.
• W2893822766 hasConcept C108583219 @default.
• W2893822766 hasConcept C111919701 @default.
• W2893822766 hasConcept C119857082 @default.
• W2893822766 hasConcept C127413603 @default.

• next