Facial Attribute Recognition (FAR) holds significant potential for wide-ranging applications. However, traditionally trained FAR models exhibit unfairness, largely due to data bias—where certain sensitive attributes correlate statistically with target attributes. To address this, we propose a group-attention mechanism: first, each image is categorized into subgroups (e.g., Male/Female&short hair, Male/Female&long hair). Within the attention mechanism, distinct Query parameters are used for each group, with shared Key and Value parameters. As group-specific Query parameters are trained on subgrouped data, the noted bias is effectively mitigated. Consequently, integrating this Group-Attention into Vision Transformer (ViT) yields our novel Group-Decoupled ViT (GD-ViT) model. Moreover, to further attenuate the statistical correlation between sensitive and target attributes, we propose a Mask-Guided Correlation Suppression learning strategy. Specifically, in Stage 1, it first leverages a min-max dual-loss optimization strategy to train GD-ViT in capturing key regions related to sensitive attributes yet irrelevant to target attributes. Then, in Stage 2, it trains another GD-ViT by masking sensitive regions identified in Stage 1, fusing the masked output (as intermediate input) with the model’s intermediate outputs. This weakens regions associated with sensitive attributes while enhancing others, suppressing the learning of key features related to sensitive attributes. Consequently, it encourages the model to focus more on intrinsic target attribute regions and balances the learning process between the sensitive attribute and the target attribute. Extensive experiments demonstrate that our method achieves superior performance across three benchmark datasets for fair facial attribute recognition.

In continual learning (CL), the newly arrived data are often out-of-distribution from the previous ones, causing drastic representation shift (RS) when updating the old model on the new data, leading to catastrophic forgetting. In this work, we propose feature boosting calibration (FBC) to tackle this problem. Specifically, an expanded module is trained to learn all the classes, including the old and new classes, discovering critical features missed by the original/old model. Then, an FBC network (FBCN) is trained to exploit these missed features to calibrate the old representations. As the missed features increase the information needed for distinguishing between the old and new classes, FBCN generates the calibrated ones with more transferable features, thus alleviating the RS. Next, given the limited memory to store samples of the old/learned classes, the data are severely imbalanced between the old and new classes. To cope with this problem, we propose blockwise knowledge distillation (BWKD), which splits the softmax layer into blocks according to class frequency and then distills each block separately, resolving data imbalance effectively. Building upon the two improvements, we propose a two-stage training framework for CL, named CKDF-V2, providing an enhanced version of the cascaded knowledge distillation framework (CKDF). Furthermore, we integrate it with a task-token expansion method to develop a novel approach for CL based on the vision transformer (ViT). Extensive experiments show that both a convolutional neural network (CNN) and ViT-based CKDF-V2 obtain favorable results across multiple CL benchmarks.

Rehearsal methods based on knowledge distillation (KD) have been widely used in continual learning (CL). However, given memory constraints, few exemplars contain limited variations of previously learned tasks, impeding the effectiveness of KD in retaining long-term knowledge. The decision boundaries learned by the typical KD strategy overfit the limited exemplars, leading to “shrunk boundaries" of the old classes. To tackle this problem, we propose a novel KD strategy, called One-to-Many Information Matching method (O2MIM), which generates interpolated data by mixing samples between old and new classes, disentangles the supervision information from them and assigns supervision information to them in favor of the old classes. By doing so, the supervision information from a single exemplar can be matched with multiple information from different interpolated images. Moreover, O2MIM utilizes one trainable parameter to create an adaptive KD loss, thereby facilitating a flexible matching process with the designated supervision information. Consequently, O2MIM exploits the exemplar corset more effectively, expanding the shrunk decision boundaries towards the new classes. Next, to incorporate new classes into our classification model, we apply an effective classification training strategy to train a debiased classifier. Combining it with O2MIM, we propose the method of Expanding the Shrinking Decision Boundaries (ESDB), which simultaneously transfers knowledge from the old model via O2MIM and learns new classes by the classification training strategy. Extensive experiments demonstrate that ESDB achieves state-of-the-art performance on diverse CL benchmarks. We also confirm that O2MIM can be used with various label-mixing methods to improve overall performance in CL. The code is available at: https://github.com/CSTiger77/ESDB.

Recently, owing to the superior performances, knowledge distillation-based (kd-based) methods with the exemplar rehearsal have been widely applied in class incremental learning (CIL). However, we discover that they suffer from the feature uncalibration problem, which is caused by directly transferring knowledge from the old model immediately to the new model when learning a new task. As the old model confuses the feature representations between the learned and new classes, the kd loss and the classification loss used in kd-based methods are heterogeneous. This is detrimental if we learn the existing knowledge from the old model directly in the way as in typical kd-based methods. To tackle this problem, the feature calibration network (FCN) is proposed, which is used to calibrate the existing knowledge to alleviate the feature representation confusion of the old model. In addition, to relieve the task-recency bias of FCN caused by the limited storage memory in CIL, we propose a novel image-feature hybrid sample rehearsal strategy to train FCN by splitting the memory budget to store the image-and-feature exemplars of the previous tasks. As feature embeddings of images have much lower-dimensions, this allows us to store more samples to train FCN. Based on these two improvements, we propose the Cascaded Knowledge Distillation Framework (CKDF) including three main stages. The first stage is used to train FCN to calibrate the existing knowledge of the old model. Then, the new model is trained simultaneously by transferring knowledge from the calibrated teacher model through the knowledge distillation strategy and learning new classes. Finally, after completing the new task learning, the feature exemplars of previous tasks are updated. Importantly, we demonstrate that the proposed CKDF is a general framework that can be applied to various kd-based methods. Experimental results show that our method achieves state-of-the-art performances on several CIL benchmarks.