Vision-language pre-training methods, e.g., CLIP, demonstrate an impressive zero-shot performance on visual categorizations with the class proxy from the text embedding of the class name.

Vision-language pre-training methods, e.g., CLIP, demonstrate an impressive zero-shot performance on visual categorizations with the class proxy from the text embedding of the class name. However, the modality gap between the text and vision space can result in a sub-optimal performance. We theoretically show that the gap cannot be reduced sufficiently by minimizing the contrastive loss in CLIP and the optimal proxy for vision tasks may reside only in the vision space. Therefore, given unlabeled target vision data, we propose to learn the vision proxy directly with the help from the text proxy for zero-shot transfer. Moreover, according to our theoretical analysis, strategies are developed to further refine the pseudo label obtained by the text proxy to facilitate the intra-modal proxy learning (InMaP) for vision. Experiments on extensive downstream tasks confirm the effectiveness and efficiency of our proposal. Concretely, InMaP can obtain the vision proxy within one minute on a single GPU while improving the zero-shot accuracy from $77.02\%$ to $80.21\%$ on ImageNet with ViT-L/14@336 pre-trained by CLIP.

Vision-language pre-training methods, e.g., CLIP, demonstrate an impressive zero-shot performance on visual categorizations with the class proxy from the text embedding of the class name.

Vision-language pre-training methods, e.g., CLIP, demonstrate an impressive zero-shot performance on visual categorizations with the class proxy from the text embedding of the class name. However, the modality gap between the text and vision space can result in a sub-optimal performance. We theoretically show that the gap cannot be reduced sufficiently by minimizing the contrastive loss in CLIP and the optimal proxy for vision tasks may reside only in the vision space. Therefore, given unlabeled target vision data, we propose to learn the vision proxy directly with the help from the text proxy for zero-shot transfer. Moreover, according to our theoretical analysis, strategies are developed to further refine the pseudo label obtained by the text proxy to facilitate the intra-modal proxy learning (InMaP) for vision.

The application of Contrastive Language-Image Pre-training (CLIP) in Weakly Supervised Semantic Segmentation (WSSS) research powerful cross-modal semantic understanding capabilities. Existing methods attempt to optimize input text prompts for improved alignment of images and text, by finely adjusting text prototypes to facilitate semantic matching. Nevertheless, given the modality gap between text and vision spaces, the text prototypes employed by these methods have not effectively established a close correspondence with pixel-level vision features. In this work, our theoretical analysis indicates that the inherent modality gap results in misalignment of text and region features, and that this gap cannot be sufficiently reduced by minimizing contrast loss in CLIP. To mitigate the impact of the modality gap, we propose a Vision Prototype Learning (VPL) framework, by introducing more representative vision prototypes. The core of this framework is to learn class-specific vision prototypes in vision space with the help of text prototypes, for capturing high-quality localization maps. Moreover, we propose a regional semantic contrast module that contrasts regions embedding with corresponding prototypes, leading to more comprehensive and robust feature learning. Experimental results show that our proposed framework achieves state-of-the-art performance on two benchmark datasets.

Vision-language pre-training such as CLIP enables zero-shot transfer that can classify images according to the candidate class names. While CLIP demonstrates an impressive zero-shot performance on diverse downstream tasks, the distribution from the target data has not been leveraged sufficiently. In this work, we study a novel online zero-shot transfer scenario, where each image arrives in a random order for classification and is visited only once to obtain prediction immediately without storing its representation. Compared with the vanilla zero-shot classification, the proposed framework preserves its flexibility for online service while considering the statistics of the arrived images as the side information to capture the distribution of target data, which can help improve the performance of real-world applications. To tackle the challenge of effective online optimization, we first develop online label learning to model the target data distribution. Then, the proxy of each class in the vision space is further optimized with the proposed online proxy learning method to mitigate the modality gap between images and text. The convergence of both online strategies can be theoretically guaranteed. By combining the predicted label from the online label learning and proxy learning, our online zero-shot transfer method (OnZeta) achieves $78.94\%$ accuracy on ImageNet without accessing the entire data set. Moreover, extensive experiments on other 13 downstream tasks with different vision encoders show a more than $3\%$ improvement on average, which demonstrates the effectiveness of our proposal. Code is available at \url{https://github.com/idstcv/OnZeta}.

Vision-language pre-training methods, e.g., CLIP, demonstrate an impressive zero-shot performance on visual categorizations with the class proxy from the text embedding of the class name. However, the modality gap between the text and vision space can result in a sub-optimal performance. We theoretically show that the gap cannot be reduced sufficiently by minimizing the contrastive loss in CLIP and the optimal proxy for vision tasks may reside only in the vision space. Therefore, given unlabeled target vision data, we propose to learn the vision proxy directly with the help from the text proxy for zero-shot transfer. Moreover, according to our theoretical analysis, strategies are developed to further refine the pseudo label obtained by the text proxy to facilitate the intra-modal proxy learning (InMaP) for vision. Experiments on extensive downstream tasks confirm the effectiveness and efficiency of our proposal. Concretely, InMaP can obtain the vision proxy within one minute on a single GPU while improving the zero-shot accuracy from $77.02\%$ to $80.21\%$ on ImageNet with ViT-L/14@336 pre-trained by CLIP. Code is available at \url{https://github.com/idstcv/InMaP}.

There is invariably a trade-off between safety and efficiency for collaborative robots (cobots) in human-robot collaborations. Robots that interact minimally with humans can work with high speed and accuracy but cannot adapt to new tasks or respond to unforeseen changes, whereas robots that work closely with humans can but only by becoming passive to humans, meaning that their main tasks suspended and efficiency compromised. Accordingly, this paper proposes a new complementary framework for human-robot collaboration that balances the safety of humans and the efficiency of robots. In this framework, the robot carries out given tasks using a vision-based adaptive controller, and the human expert collaborates with the robot in the null space. Such a decoupling drives the robot to deal with existing issues in task space (e.g., uncalibrated camera, limited field of view) and in null space (e.g., joint limits) by itself while allowing the expert to adjust the configuration of the robot body to respond to unforeseen changes (e.g., sudden invasion, change of environment) without affecting the robot's main task. Additionally, the robot can simultaneously learn the expert's demonstration in task space and null space beforehand with dynamic movement primitives (DMP). Therefore, an expert's knowledge and a robot's capability are both explored and complementary. Human demonstration and involvement are enabled via a mixed interaction interface, i.e., augmented reality (AR) and haptic devices. The stability of the closed-loop system is rigorously proved with Lyapunov methods. Experimental results in various scenarios are presented to illustrate the performance of the proposed method.