Mainstream unsupervised domain adaptation (UDA) methods align feature distributions across different domains via adversarial learning. However, most of them focus on global distribution alignment, ignoring the fine-grained domain discrepancy. Besides, they generally require auxiliary models, bringing extra computation costs. To tackle these issues, this study proposes an UDA method that differentiates individual samples without the help of extra models. To this end, we introduce a novel discrepancy metric, termed style discrepancy, to distinguish different target samples. We also propose a paradigm for adversarial style discrepancy minimization (ASDM). Specifically, we fix the parameters of the feature extractor and maximize style discrepancy to update the classifier, which helps detect more hard samples. Adversely, we fix the parameters of the classifier and minimize the style discrepancy to update the feature extractor, pushing those hard samples near the support of the source distribution. Such adversary helps to progressively detect and adapt more hard samples, leading to fine-grained domain adaptation. Experiments on different UDA tasks validate the effectiveness of ASDM. Overall, without any extra models, ASDM reaches a 46.9% mIoU in the GTA5 to Cityscapes benchmark and an 84.7% accuracy in the VisDA-2017 benchmark, outperforming many existing adversarial-learning-based methods.

Self-supervised pre-training has become the priory choice to establish reliable neural networks for automated recognition of massive biomedical microscopy images, which are routinely annotation-free, without semantics, and without guarantee of quality. Note that this paradigm is still at its infancy and limited by closely related open issues: (1) how to learn robust representations in an unsupervised manner from unlabeled biomedical microscopy images of low diversity in samples? and (2) how to obtain the most significant representations demanded by a high-quality segmentation? Aiming at these issues, this study proposes a knowledge-based learning framework (TOWER) towards enhanced recognition of biomedical microscopy images, which works in three phases by synergizing contrastive learning and generative learning methods: (1) Sample Space Diversification: Reconstructive proxy tasks have been enabled to embed a priori knowledge with context highlighted to diversify the expanded sample space; (2) Enhanced Representation Learning: Informative noise-contrastive estimation loss regularizes the encoder to enhance representation learning of annotation-free images; (3) Correlated Optimization: Optimization operations in pre-training the encoder and the decoder have been correlated via image restoration from proxy tasks, targeting the need for semantic segmentation. Experiments have been conducted on public datasets of biomedical microscopy images against the state-of-the-art counterparts (e.g., SimCLR and BYOL), and results demonstrate that: TOWER statistically excels in all self-supervised methods, achieving a Dice improvement of 1.38 percentage points over SimCLR. TOWER also has potential in multi-modality medical image analysis and enables label-efficient semi-supervised learning, e.g., reducing the annotation cost by up to 99% in pathological classification.

Protocol Reverse Engineering (PRE) is crucial for information security of Internet-of-Things (IoT), and message clustering determines the effectiveness of PRE. However, the quality of services still lags behind the strict requirement of IoT applications as the results of message clustering are often coarse-grained with the intrinsic type information hidden in messages largely ignored. Aiming at this problem, this study proposes a type-aware approach to message clustering guided by type information. The approach regards a message as a combination of n-grams, and it employs the Latent Dirichlet Allocation (LDA) model to characterize messages with types and n-grams via inferring the type distribution of each message. The type distribution is finally used to measure the similarity of messages. According to this similarity, the approach clusters messages and further extracts message formats. Experimental results of the approach against Netzob in terms of a number of protocols indicate that the correctness and conciseness can be significantly improved, e.g., figures 43.86% and 3.87%, respectively for the CoAP protocol.

Stereo matching is an important research field of computer vision. Due to the dimension of cost aggregation, current neural network-based stereo methods are difficult to trade-off speed and accuracy. To this end, we integrate fast 2D stereo methods with accurate 3D networks to improve performance and reduce running time. We leverage a 2D encoder-decoder network to generate a rough disparity map and construct a disparity range to guide the 3D aggregation network, which can significantly improve the accuracy and reduce the computational cost. We use a stacked hourglass structure to refine the disparity from coarse to fine. We evaluated our method on three public datasets. According to the KITTI official website results, Our network can generate an accurate result in 80 ms on a modern GPU. Compared to other 2D stereo networks (AANet, DeepPruner, FADNet, etc.), our network has a big improvement in accuracy. Meanwhile, it is significantly faster than other 3D stereo networks (5× than PSMNet, 7.5× than CSN and 22.5× than GANet, etc.), demonstrating the effectiveness of our method.

Source-free model adaptation (SFMA) plays an important role in robust robot automation, which aims to mitigate distributional inconsistency between source and target data while avoiding accessing source data. SFMA methods generally benefit from self-training, which can be prone to overfitting noisy pseudo labels. Previous methods used various techniques to filter out noisy predictions and improve self-training performance. Nonetheless, the filtering is heavily reliant on pre-defined thresholds and suffers from a class imbalance problem, resulting in a bias towards the majority classes and poor performance of the other classes. Aiming at this pitfall, this study proposes to combat the noisy pseudo labels via multi-level perturbation, which involves all the pseudo predictions in self-training. The proposed method introduces different perturbations at three different levels: input, feature, and model. These perturbations boost the diversity of training data and increase the difficulty of fitting pseudo labels, which avoids overfitting the noise without ignoring the minority classes. Besides, the proposed method helps mitigate the intra-domain gap by maximizing the consistency between the original predictions and the perturbed predictions. Experimental results on two domain adaptive segmentation benchmarks confirm the effectiveness of the proposed method, which outperforms state-of-the-art SFMA methods.