| Full text | |
| Author(s): |
Alvarenga e Silva, Lucas Fernando
;
Sebe, Nicu
;
Almeida, Jurandy
Total Authors: 3
|
| Document type: | Journal article |
| Source: | 2023 36TH CONFERENCE ON GRAPHICS, PATTERNS AND IMAGES, SIBGRAPI 2023; v. N/A, p. 6-pg., 2023-01-01. |
| Abstract | |
Convolutional Neural Networks (CNNs) have brought revolutionary advances to many research areas due to their capacity of learning from raw data. However, when those methods are applied to non-controllable environments, many different factors can degrade the model's expected performance, such as unlabeled datasets with different levels of domain shift and category shift. Particularly, when both issues occur at the same time, we tackle this challenging setup as Open Set Domain Adaptation (OSDA) problem. In general, existing OSDA approaches focus their efforts only on aligning known classes or, if they already extract possible negative instances, use them as a new category learned with supervision during the course of training. We propose a novel way to improve OSDA approaches by extracting a high-confidence set of unknown instances and using it as a hard constraint to tighten the classification boundaries of OSDA methods. Especially, we adopt a new loss constraint evaluated in three different means, (1) directly with the pristine negative instances; (2) with randomly transformed negatives using data augmentation techniques; and (3) with synthetically generated negatives containing adversarial features. We assessed all approaches in an extensive set of experiments based on OVANet, where we could observe consistent improvements for two public benchmarks, the Office-31 and Office-Home datasets, yielding absolute gains of up to 1.3% for both Accuracy and H-Score on Office-31 and 5.8% for Accuracy and 4.7% for H-Score on Office-Home. (AU) | |
| FAPESP's process: | 17/25908-6 - Weakly supervised learning for compressed video analysis on retrieval and classification tasks for visual alert |
| Grantee: | João Paulo Papa |
| Support Opportunities: | Research Grants - Research Partnership for Technological Innovation - PITE |
| FAPESP's process: | 23/03328-9 - Machine Learning Under The Hood: Efficient Accelerators for Deep Networks and its Applicability in Scientific Computation |
| Grantee: | Lucas Fernando Alvarenga e Silva |
| Support Opportunities: | Scholarships in Brazil - Doctorate |
| FAPESP's process: | 20/08770-3 - Open set methods based on deep networks for multimedia recognition |
| Grantee: | Lucas Fernando Alvarenga e Silva |
| Support Opportunities: | Scholarships in Brazil - Master |
| FAPESP's process: | 21/13348-1 - Investigation of open set domain adaptation methods for computer vision tasks |
| Grantee: | Lucas Fernando Alvarenga e Silva |
| Support Opportunities: | Scholarships abroad - Research Internship - Master's degree |