Color Constancy Convolutional Autoencoder
Tutkimustuotos › › vertaisarvioitu
Yksityiskohdat
| Alkuperäiskieli | Englanti |
|---|---|
| Otsikko | 2019 IEEE Symposium Series on Computational Intelligence, SSCI 2019 |
| Kustantaja | IEEE |
| Sivut | 1085-1090 |
| Sivumäärä | 6 |
| ISBN (elektroninen) | 9781728124858 |
| ISBN (painettu) | 978-1-7281-2486-5 |
| DOI - pysyväislinkit | |
| Tila | Julkaistu - 2019 |
| OKM-julkaisutyyppi | A4 Artikkeli konferenssijulkaisussa |
| Tapahtuma | IEEE Symposium Series on Computational Intelligence - Kesto: 1 tammikuuta 1900 → … |
Conference
| Conference | IEEE Symposium Series on Computational Intelligence |
|---|---|
| Lyhennettä | IEEE SSCI |
| Ajanjakso | 1/01/00 → … |
Tiivistelmä
In this paper, we study the importance of pretraining for the generalization capability in the color constancy problem. We propose two novel approaches based on convolutional autoencoders: an unsupervised pre-training algorithm using a fine-tuned encoder and a semi-supervised pre-training algorithm using a novel composite-loss function. This enables us to solve the data scarcity problem and achieve competitive, to the state-of-the-art, results while requiring much fewer parameters on ColorChecker RECommended dataset. We further study the over-fitting phenomenon on the recently introduced version of INTEL-TUT Dataset for Camera Invariant Color Constancy Research, which has both field and non-field scenes acquired by three different camera models.