CDAD-Net: Bridging Domain Gaps in Generalized Category Discovery
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866913305414598656 |
|---|---|
| author | Rongali, Sai Bhargav Mehrotra, Sarthak Jha, Ankit C, Mohamad Hassan N Bose, Shirsha Gupta, Tanisha Singha, Mainak Banerjee, Biplab |
| author_facet | Rongali, Sai Bhargav Mehrotra, Sarthak Jha, Ankit C, Mohamad Hassan N Bose, Shirsha Gupta, Tanisha Singha, Mainak Banerjee, Biplab |
| contents | In Generalized Category Discovery (GCD), we cluster unlabeled samples of known and novel classes, leveraging a training dataset of known classes. A salient challenge arises due to domain shifts between these datasets. To address this, we present a novel setting: Across Domain Generalized Category Discovery (AD-GCD) and bring forth CDAD-NET (Class Discoverer Across Domains) as a remedy. CDAD-NET is architected to synchronize potential known class samples across both the labeled (source) and unlabeled (target) datasets, while emphasizing the distinct categorization of the target data. To facilitate this, we propose an entropy-driven adversarial learning strategy that accounts for the distance distributions of target samples relative to source-domain class prototypes. Parallelly, the discriminative nature of the shared space is upheld through a fusion of three metric learning objectives. In the source domain, our focus is on refining the proximity between samples and their affiliated class prototypes, while in the target domain, we integrate a neighborhood-centric contrastive learning mechanism, enriched with an adept neighborsmining approach. To further accentuate the nuanced feature interrelation among semantically aligned images, we champion the concept of conditional image inpainting, underscoring the premise that semantically analogous images prove more efficacious to the task than their disjointed counterparts. Experimentally, CDAD-NET eclipses existing literature with a performance increment of 8-15% on three AD-GCD benchmarks we present. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_05366 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | CDAD-Net: Bridging Domain Gaps in Generalized Category Discovery Rongali, Sai Bhargav Mehrotra, Sarthak Jha, Ankit C, Mohamad Hassan N Bose, Shirsha Gupta, Tanisha Singha, Mainak Banerjee, Biplab Computer Vision and Pattern Recognition In Generalized Category Discovery (GCD), we cluster unlabeled samples of known and novel classes, leveraging a training dataset of known classes. A salient challenge arises due to domain shifts between these datasets. To address this, we present a novel setting: Across Domain Generalized Category Discovery (AD-GCD) and bring forth CDAD-NET (Class Discoverer Across Domains) as a remedy. CDAD-NET is architected to synchronize potential known class samples across both the labeled (source) and unlabeled (target) datasets, while emphasizing the distinct categorization of the target data. To facilitate this, we propose an entropy-driven adversarial learning strategy that accounts for the distance distributions of target samples relative to source-domain class prototypes. Parallelly, the discriminative nature of the shared space is upheld through a fusion of three metric learning objectives. In the source domain, our focus is on refining the proximity between samples and their affiliated class prototypes, while in the target domain, we integrate a neighborhood-centric contrastive learning mechanism, enriched with an adept neighborsmining approach. To further accentuate the nuanced feature interrelation among semantically aligned images, we champion the concept of conditional image inpainting, underscoring the premise that semantically analogous images prove more efficacious to the task than their disjointed counterparts. Experimentally, CDAD-NET eclipses existing literature with a performance increment of 8-15% on three AD-GCD benchmarks we present. |
| title | CDAD-Net: Bridging Domain Gaps in Generalized Category Discovery |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2404.05366 |