We lifted 13 functions out of this paper's own repositories and ran 10 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| tim-learn/Dis-tune | canonical | 6 of 7 |
| tim-learn/SHOT | canonical | 3 of 5 |
| tim-learn/dine | canonical | 1 of 1 |
| Function | Status | Where it lives |
|---|---|---|
| Entropy | Ran | tim-learn/Dis-tune/loss.py code served (permissive licence) · get_code("94b5622f0aa7add1") |
| Entropy | Ran | tim-learn/SHOT/digit/loss.py code served (permissive licence) · get_code("866722884763b224") |
| grl_hook | Ran | tim-learn/SHOT/object/loss.py code served (permissive licence) · get_code("9768efb52f591b55") |
| image_test | Ran | tim-learn/SHOT/object/image_multisource.py code served (permissive licence) · get_code("6e4265b7f707b718") |
| image_train | Ran | tim-learn/Dis-tune/DINE_dist.py code served (permissive licence) · get_code("35c2d572ebd185f5") |
| l_loader | Ran | tim-learn/Dis-tune/data_list.py code served (permissive licence) · get_code("edd7184ac144c4fa") |
| lr_scheduler | Ran | tim-learn/Dis-tune/DINE_dist.py code served (permissive licence) · get_code("0b7ffc9f8b77529c") |
| lr_scheduler | Ran | tim-learn/dine/DINE_ft.py code served (permissive licence) · get_code("8bd136d4257e40ea") |
| op_copy | Ran | tim-learn/Dis-tune/DINE_dist.py code served (permissive licence) · get_code("93a11f62e4a129f0") |
| rgb_loader | Ran | tim-learn/Dis-tune/data_list.py code served (permissive licence) · get_code("2c5ce24ea2b5d2a4") |
| CDAN | Not yet run | tim-learn/SHOT/object/loss.py code served (permissive licence) · get_code("385d1e010dfb76a9") |
| calc_coeff | Not yet run | tim-learn/SHOT/object/network.py code served (permissive licence) · get_code("e352afa5762c5a6b") |
| make_dataset | Not yet run | tim-learn/Dis-tune/data_list.py code served (permissive licence) · get_code("2301055cb33836bc") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
To ease the burden of labeling, unsupervised domain adaptation (UDA) aims to transfer knowledge in previous and related labeled datasets (sources) to a new unlabeled dataset (target). Despite impressive progress, prior methods always need to access the raw source data and develop data-dependent alignment approaches to recognize the target samples in a transductive learning manner, which may raise privacy concerns from source individuals. Several recent studies resort to an alternative solution by exploiting the well-trained white-box model from the source domain, yet, it may still leak the raw data through generative adversarial learning. This paper studies a practical and interesting setting for UDA, where only black-box source models (i.e., only network predictions are available) are provided during adaptation in the target domain. To solve this problem, we propose a new two-step knowledge adaptation framework called DIstill and fine-tuNE (DINE). Taking into consideration the target data structure, DINE first distills the knowledge from the source predictor to a customized target model, then fine-tunes the distilled model to further fit the target domain. Besides, neural networks are not required to be identical across domains in DINE, even allowing effective adaptation on a low-resource device. Empirical results on three UDA scenarios (i.e., single-source, multi-source, and partial-set) confirm that DINE achieves highly competitive performance compared to state-of-the-art data-dependent approaches. Code is available at \url{https://github.com/tim-learn/DINE/}.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2104.01539")
get_code_for_paper("2104.01539")
have("2104.01539")
Connect an agent — have() is free.