The Devil is in the Few Shots: Iterative Visual Knowledge Completion for Few-shot Learning

Contrastive Language-Image Pre-training (CLIP) has shown powerful zero-shot learning performance. Few-shot learning aims to further enhance the transfer capability of CLIP by giving few images in each class, aka 'few shots'. Most existing methods either implicitly learn from the few shots by incorporating learnable prompts or adapters, or explicitly embed them in a cache model for inference. However, the narrow distribution of few shots often contains incomplete class information, leading to biased visual knowledge with high risk of misclassification. To tackle this problem, recent methods propose to supplement visual knowledge by generative models or extra databases, which can be costly and time-consuming. In this paper, we propose an Iterative Visual Knowledge CompLetion (KCL) method to complement visual knowledge by properly taking advantages of unlabeled samples without access to any auxiliary or synthetic data. Specifically, KCL first measures the similarities between unlabeled samples and each category. Then, the samples with top confidence to each category is selected and collected by a designed confidence criterion. Finally, the collected samples are treated as labeled ones and added to few shots to jointly re-estimate the remaining unlabeled ones. The above procedures will be repeated for a certain number of iterations with more and more samples being collected until convergence, ensuring a progressive and robust knowledge completion process. Extensive experiments on 11 benchmark datasets demonstrate the effectiveness and efficiency of KCL as a plug-and-play module under both few-shot and zero-shot learning settings. Code is available at https://github.com/Mark-Sky/KCL.

Conditional Prototype Rectification Prompt Learning

Pre-trained large-scale vision-language models (VLMs) have acquired profound understanding of general visual concepts. Recent advancements in efficient transfer learning (ETL) have shown remarkable success in fine-tuning VLMs within the scenario of limited data, introducing only a few parameters to harness task-specific insights from VLMs. Despite significant progress, current leading ETL methods tend to overfit the narrow distributions of base classes seen during training and encounter two primary challenges: (i) only utilizing uni-modal information to modeling task-specific knowledge; and (ii) using costly and time-consuming methods to supplement knowledge. To address these issues, we propose a Conditional Prototype Rectification Prompt Learning (CPR) method to correct the bias of base examples and augment limited data in an effective way. Specifically, we alleviate overfitting on base classes from two aspects. First, each input image acquires knowledge from both textual and visual prototypes, and then generates sample-conditional text tokens. Second, we extract utilizable knowledge from unlabeled data to further refine the prototypes. These two strategies mitigate biases stemming from base classes, yielding a more effective classifier. Extensive experiments on 11 benchmark datasets show that our CPR achieves state-of-the-art performance on both few-shot classification and base-to-new generalization tasks. Our code is avaliable at \url{https://github.com/chenhaoxing/CPR}.

Segment Anything Model Meets Image Harmonization

Image harmonization is a crucial technique in image composition that aims to seamlessly match the background by adjusting the foreground of composite images. Current methods adopt either global-level or pixel-level feature matching. Global-level feature matching ignores the proximity prior, treating foreground and background as separate entities. On the other hand, pixel-level feature matching loses contextual information. Therefore, it is necessary to use the information from semantic maps that describe different objects to guide harmonization. In this paper, we propose Semantic-guided Region-aware Instance Normalization (SRIN) that can utilize the semantic segmentation maps output by a pre-trained Segment Anything Model (SAM) to guide the visual consistency learning of foreground and background features. Abundant experiments demonstrate the superiority of our method for image harmonization over state-of-the-art methods.

Boosting Audio-visual Zero-shot Learning with Large Language Models

Audio-visual zero-shot learning aims to recognize unseen categories based on paired audio-visual sequences. Recent methods mainly focus on learning aligned and discriminative multi-modal features to boost generalization towards unseen categories. However, these approaches ignore the obscure action concepts in category names and may inevitably introduce complex network structures with difficult training objectives. In this paper, we propose a simple yet effective framework named Knowledge-aware Distribution Adaptation (KDA) to help the model better grasp the novel action contents with an external knowledge base. Specifically, we first propose using large language models to generate rich descriptions from category names, which leads to a better understanding of unseen categories. Additionally, we propose a distribution alignment loss as well as a knowledge-aware adaptive margin loss to further improve the generalization ability towards unseen categories. Extensive experimental results demonstrate that our proposed KDA can outperform state-of-the-art methods on three popular audio-visual zero-shot learning datasets. Our code will be avaliable at \url{https://github.com/chenhaoxing/KDA}.

DiffUTE: Universal Text Editing Diffusion Model

Diffusion model based language-guided image editing has achieved great success recently. However, existing state-of-the-art diffusion models struggle with rendering correct text and text style during generation. To tackle this problem, we propose a universal self-supervised text editing diffusion model (DiffUTE), which aims to replace or modify words in the source image with another one while maintaining its realistic appearance. Specifically, we build our model on a diffusion model and carefully modify the network structure to enable the model for drawing multilingual characters with the help of glyph and position information. Moreover, we design a self-supervised learning framework to leverage large amounts of web data to improve the representation ability of the model. Experimental results show that our method achieves an impressive performance and enables controllable editing on in-the-wild images with high fidelity. Our code will be avaliable in \url{https://github.com/chenhaoxing/DiffUTE}.

Mobile User Interface Element Detection Via Adaptively Prompt Tuning

Recent object detection approaches rely on pretrained vision-language models for image-text alignment. However, they fail to detect the Mobile User Interface (MUI) element since it contains additional OCR information, which describes its content and function but is often ignored. In this paper, we develop a new MUI element detection dataset named MUI-zh and propose an Adaptively Prompt Tuning (APT) module to take advantage of discriminating OCR information. APT is a lightweight and effective module to jointly optimize category prompts across different modalities. For every element, APT uniformly encodes its visual features and OCR descriptions to dynamically adjust the representation of frozen category prompts. We evaluate the effectiveness of our plug-and-play APT upon several existing CLIP-based detectors for both standard and open-vocabulary MUI element detection. Extensive experiments show that our method achieves considerable improvements on two datasets. The datasets is available at \url{github.com/antmachineintelligence/MUI-zh}.

DiffusionInst: Diffusion Model for Instance Segmentation

Recently, diffusion frameworks have achieved comparable performance with previous state-of-the-art image generation models. Researchers are curious about its variants in discriminative tasks because of its powerful noise-to-image denoising pipeline. This paper proposes DiffusionInst, a novel framework that represents instances as instance-aware filters and formulates instance segmentation as a noise-to-filter denoising process. The model is trained to reverse the noisy groundtruth without any inductive bias from RPN. During inference, it takes a randomly generated filter as input and outputs mask in one-step or multi-step denoising. Extensive experimental results on COCO and LVIS show that DiffusionInst achieves competitive performance compared to existing instance segmentation models. We hope our work could serve as a simple yet effective baseline, which could inspire designing more efficient diffusion frameworks for challenging discriminative tasks. Our code is available in https://github.com/chenhaoxing/DiffusionInst.

Hierarchical Dynamic Image Harmonization

Image harmonization is a critical task in computer vision, which aims to adjust the fore-ground to make it compatible with the back-ground. Recent works mainly focus on using global transformation (i.e., normalization and color curve rendering) to achieve visual consistency. However, these model ignore local consistency and their model size limit their harmonization ability on edge devices. Inspired by the dynamic deep networks that adapt the model structures or parameters conditioned on the inputs, we propose a hierarchical dynamic network (HDNet) for efficient image harmonization to adapt the model parameters and features from local to global view for better feature transformation. Specifically, local dynamics (LD) and mask-aware global dynamics (MGD) are applied. LD enables features of different channels and positions to change adaptively and improve the representation ability of geometric transformation through structural information learning. MGD learns the representations of fore- and back-ground regions and correlations to global harmonization. Experiments show that the proposed HDNet reduces more than 80\% parameters compared with previous methods but still achieves the state-of-the-art performance on the popular iHarmony4 dataset. Our code is avaliable in https://github.com/chenhaoxing/HDNet.

Shaping Visual Representations with Attributes for Few-Shot Learning

Few-shot recognition aims to recognize novel categories under low-data regimes. Due to the scarcity of images, machines cannot obtain enough effective information, and the generalization ability of the model is extremely weak. By using auxiliary semantic modalities, recent metric-learning based few-shot learning methods have achieved promising performances. However, these methods only augment the representations of support classes, while query images have no semantic modalities information to enhance representations. Instead, we propose attribute-shaped learning (ASL), which can normalize visual representations to predict attributes for query images. And we further devise an attribute-visual attention module (AVAM), which utilizes attributes to generate more discriminative features. Our method enables visual representations to focus on important regions with attributes guidance. Experiments demonstrate that our method can achieve competitive results on CUB and SUN benchmarks. Our code is available at {https://github.com/chenhaoxing/ASL}.

Sparse Spatial Transformers for Few-Shot Learning

Learning from limited data is a challenging task since the scarcity of data leads to a poor generalization of the trained model. The classical global pooled representation is likely to lose useful local information. Recently, many few shot learning methods address this challenge by using deep descriptors and learning a pixel-level metric. However, using deep descriptors as feature representations may lose the contextual information of the image. And most of these methods deal with each class in the support set independently, which cannot sufficiently utilize discriminative information and task-specific embeddings. In this paper, we propose a novel Transformer based neural network architecture called Sparse Spatial Transformers (SSFormers), which can find task-relevant features and suppress task-irrelevant features. Specifically, we first divide each input image into several image patches of different sizes to obtain dense local features. These features retain contextual information while expressing local information. Then, a sparse spatial transformer layer is proposed to find spatial correspondence between the query image and the entire support set to select task-relevant image patches and suppress task-irrelevant image patches. Finally, we propose an image patch matching module to calculate the distance between dense local representations to determine which category the query image belongs to in the support set. Extensive experiments on popular few-shot learning benchmarks show that our method achieves the state-of-the-art performance. Our code is available at \url{https://github.com/chenhaoxing/SSFormers}.