|
|
|
|
| LEADER |
01000naa a22002652 4500 |
| 001 |
NLM333507142 |
| 003 |
DE-627 |
| 005 |
20231225221805.0 |
| 007 |
cr uuu---uuuuu |
| 008 |
231225s2022 xx |||||o 00| ||eng c |
| 024 |
7 |
|
|a 10.1109/TPAMI.2021.3129870
|2 doi
|
| 028 |
5 |
2 |
|a pubmed24n1111.xml
|
| 035 |
|
|
|a (DE-627)NLM333507142
|
| 035 |
|
|
|a (NLM)34813469
|
| 040 |
|
|
|a DE-627
|b ger
|c DE-627
|e rakwb
|
| 041 |
|
|
|a eng
|
| 100 |
1 |
|
|a Mensink, Thomas
|e verfasserin
|4 aut
|
| 245 |
1 |
0 |
|a Factors of Influence for Transfer Learning Across Diverse Appearance Domains and Task Types
|
| 264 |
|
1 |
|c 2022
|
| 336 |
|
|
|a Text
|b txt
|2 rdacontent
|
| 337 |
|
|
|a ƒaComputermedien
|b c
|2 rdamedia
|
| 338 |
|
|
|a ƒa Online-Ressource
|b cr
|2 rdacarrier
|
| 500 |
|
|
|a Date Completed 09.11.2022
|
| 500 |
|
|
|a Date Revised 19.11.2022
|
| 500 |
|
|
|a published: Print-Electronic
|
| 500 |
|
|
|a Citation Status MEDLINE
|
| 520 |
|
|
|a Transfer learning enables to re-use knowledge learned on a source task to help learning a target task. A simple form of transfer learning is common in current state-of-the-art computer vision models, i.e., pre-training a model for image classification on the ILSVRC dataset, and then fine-tune on any target task. However, previous systematic studies of transfer learning have been limited and the circumstances in which it is expected to work are not fully understood. In this paper we carry out an extensive experimental exploration of transfer learning across vastly different image domains (consumer photos, autonomous driving, aerial imagery, underwater, indoor scenes, synthetic, close-ups) and task types (semantic segmentation, object detection, depth estimation, keypoint detection). Importantly, these are all complex, structured output tasks types relevant to modern computer vision applications. In total we carry out over 2000 transfer learning experiments, including many where the source and target come from different image domains, task types, or both. We systematically analyze these experiments to understand the impact of image domain, task type, and dataset size on transfer learning performance. Our study leads to several insights and concrete recommendations: (1) for most tasks there exists a source which significantly outperforms ILSVRC'12 pre-training; (2) the image domain is the most important factor for achieving positive transfer; (3) the source dataset should include the image domain of the target dataset to achieve best results; (4) at the same time, we observe only small negative effects when the image domain of the source task is much broader than that of the target; (5) transfer across task types can be beneficial, but its success is heavily dependent on both the source and target task types
|
| 650 |
|
4 |
|a Journal Article
|
| 700 |
1 |
|
|a Uijlings, Jasper
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Kuznetsova, Alina
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Gygli, Michael
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Ferrari, Vittorio
|e verfasserin
|4 aut
|
| 773 |
0 |
8 |
|i Enthalten in
|t IEEE transactions on pattern analysis and machine intelligence
|d 1979
|g 44(2022), 12 vom: 01. Dez., Seite 9298-9314
|w (DE-627)NLM098212257
|x 1939-3539
|7 nnns
|
| 773 |
1 |
8 |
|g volume:44
|g year:2022
|g number:12
|g day:01
|g month:12
|g pages:9298-9314
|
| 856 |
4 |
0 |
|u http://dx.doi.org/10.1109/TPAMI.2021.3129870
|3 Volltext
|
| 912 |
|
|
|a GBV_USEFLAG_A
|
| 912 |
|
|
|a SYSFLAG_A
|
| 912 |
|
|
|a GBV_NLM
|
| 912 |
|
|
|a GBV_ILN_350
|
| 951 |
|
|
|a AR
|
| 952 |
|
|
|d 44
|j 2022
|e 12
|b 01
|c 12
|h 9298-9314
|