Official implementation of the paper "COMMA: Co-Articulated Multi-Modal Learning".

1. Correlated prompt generation: The prompts of the vision and language branches in these methods are usually separated or uni-directionally correlated. To better guide and align the representations of two branches, we present to compute prompts based on preceding prompts of both branches to aggregate beneficial multi-modal information.
2. Alleviating Forgetting Generic Knowledge: The essential generic knowledge learned in the pretraining stage is partly forgotten in the fine-tuning process. We propose to alleviate forgetting generic knowledge by minimizing the feature discrepancy between the learnable prompts and hand-crafted prompts of the pretrained CLIP in the last several layers.

Results reported below show accuracy for base and novel classes for across 11 recognition datasets averaged over 3 seeds.

For installation and other package requirements, please follow the instructions detailed in INSTALL.md.

Please follow the instructions at DATASETS.md to prepare all datasets.

Please refer to the RUN.md for detailed instructions on training and evaluating.

Our code is based on Co-CoOp/CoOp and MaPLe repositories. We thank the authors for releasing their code.