Although the fine-tuning-based deep transfer learning method performs well in adapting a pre-trained model to downstream tasks, it struggles in non-stationary environments where the data distribution changes dynamically over time. In such scenarios, the trained model rapidly becomes obsolete. Directly fine-tuning it for new environments will inevitably lead to negative transfer, degrading the model's generalization performance. Inspired by neuroscience, some existing studies have proposed a so-called “unlearning-and-relearning” training paradigm to alleviate this issue. However, existing methods face limitations; for example, they may destroy general features in shallow layers and lack adaptability when performing deep-layer unlearning. This paper proposes a Layer-wise Unlearning (LwU) method. Our approach first identifies the layers requiring unlearning by estimating the transferability of each layer in the trained model. Then, within the selected layers, it preserves high-sensitivity parameters and re-initializes low-sensitivity ones based on a parameter sensitivity analysis to achieve unlearning. Extensive experiments on synthetic, real-world, and image datasets demonstrate that LwU effectively overcomes negative transfer and substantially improves the model's generalization performance in new environments. The code of the proposed method and data are available at https://github.com/mlmmwym/LwU.

Multimodal Sentiment Analysis (MSA) aims to integrate textual, audio, and visual data to capture nuanced sentimental cues. Although text dominates in existing approaches, audio and visual modalities inherently contain both shared semantics (overlapping with text) and private semantics. Existing methods struggle to precisely find semantic boundaries and lack explicit mechanisms for modeling interaction between shared/private semantics and different modalities. To address this, we propose DiffuFuse, a framework that uses a diffusion denoising model to leverage textual information to predict shared semantic features, dynamically and adaptively delineate semantic boundaries for non-textual features, and employs a dual-stream fusion strategy to accurately model the interactions between different modalities and semantic types. Finally, adopt an orthogonal projection method to reduce redundancy and eliminate overlapping information between the two streams. DiffuFuse is evaluated on the MOSI and MOSEI datasets, and the experimental results demonstrate that our proposed DiffuFuse achieves superior performance.

Scene text detection is usually recognized as one of the fundamental challenges in computer vision and robotics. While the detection of scene text for English and Chinese has achieved great progress, Vietnamese scene text detection is still a relatively new field. In Vietnamese scene text detection, the diacritics of Vietnamese characters are often too small to be detected, making it difficult to accurately extract their features. The bounding boxes are usually placed close to Latin letters, resulting in incomplete covering for diacritics. To overcome these challenges, we propose a method of focusing on diacritics to detect Vietnamese scene text. Firstly, we propose detailed information embedding to enhance the features of diacritics. Secondly, we introduce IoU-v to allow the proposals to match the boundary of Vietnamese text instances more accurately. Lastly, in mask branch, we incorporate a diacritics feature fusion module to further enhance the features of diacritics. The experimental results illustrate the proposed method outperforms the existing approaches. The code of the proposed algorithm is available at https://github.com/mlmmwym/FD2VSTD.

For transfer learning, many research works have demonstrated that effective use of information from multi-source domains will improve classification performance. In this paper, we propose a method of Targetize Multi-source Domain Bridged by Common Subspace (TMSD) for face recognition, which transfers rich supervision knowledge from more than one labeled source domains to the unlabeled target domain. Specifically, a common subspace is learnt for several domains by keeping the maximum total correlation. In this way, the discrepancy of each domain is reduced, and the structures of both the source and target domains are well preserved for classification. In the common subspace, each sample projected from the source domains is sparsely represented as a linear combination of several samples projected from the target domain, such that the samples projected from different domains can be well interlaced. Then, in the original image space, each source domain image can be represented as a linear combination of neighbors in the target domain. Finally, the discriminant subspace can be obtained by targetized multi-source domain images using supervised learning algorithm. The experimental results illustrate the superiority of TMSD over those competitive ones.

Multi-label stream classification has not been fully explored for the unique properties of large data volumes, realtime, label dependencies, etc. Some methods try to take into account label dependencies, but they only focus on the existing frequent label combinations, leading to worse performance for multi-label classification. To deal with these problems, this paper proposes an algorithm which dynamically recognizes some new frequent label combinations and updates the trained classifier by class incremental learning strategy. Experimental results over both real-world and synthetic datasets demonstrate its better predictive performance.

In order to handle the information overload on the tourism websites and understand user's travel preference, we propose to build a users' preference matrix to reflect the users' preference degree on a set of scenic spots, and then propose a method of combining user clustering with Markov chain to predict the next scenic spot a user will browse on a tourism website. The experimental results indicate that the preference matrix can catch the travel preference of users and the proposed method can be used for predicting the next browsing behaviors of users.