Xin Jiang

Maofu Liu, Xin Jiang, Xiaokang Zhang

Referring Remote Sensing Image Segmentation (RRSIS) is a challenging task, aiming to segment specific target objects in remote sensing (RS) images based on a given language expression. Existing RRSIS methods typically employ coarse-grained unidirectional alignment approaches to obtain multimodal features, and they often overlook the critical role of language features as contextual information during the decoding process. Consequently, these methods exhibit weak object-level correspondence between visual and language features, leading to incomplete or erroneous predicted masks, especially when handling complex expressions and intricate RS image scenes. To address these challenges, we propose a fine-grained cross-modal alignment and decoding Transformer, CADFormer, for RRSIS. Specifically, we design a semantic mutual guidance alignment module (SMGAM) to achieve both vision-to-language and language-to-vision alignment, enabling comprehensive integration of visual and textual features for fine-grained cross-modal alignment. Furthermore, a textual-enhanced cross-modal decoder (TCMD) is introduced to incorporate language features during decoding, using refined textual information as context to enhance the relationship between cross-modal features. To thoroughly evaluate the performance of CADFormer, especially for inconspicuous targets in complex scenes, we constructed a new RRSIS dataset, called RRSIS-HR, which includes larger high-resolution RS image patches and semantically richer language expressions. Extensive experiments on the RRSIS-HR dataset and the popular RRSIS-D dataset demonstrate the effectiveness and superiority of CADFormer. Datasets and source codes will be available at https://github.com/zxk688.

Chen Chen, Xin Jiang, Weiguo Zhou et al.

In the industrial domain, the pose estimation of multiple texture-less shiny parts is a valuable but challenging task. In this particular scenario, it is impractical to utilize keypoints or other texture information because most of them are not actual features of the target but the reflections of surroundings. Moreover, the similarity of color also poses a challenge in segmentation. In this article, we propose to divide the pose estimation process into three stages: object detection, features detection and pose optimization. A convolutional neural network was utilized to perform object detection. Concerning the reliability of surface texture, we leveraged the contour information for estimating pose. Since conventional contour-based methods are inapplicable to clustered metal parts due to the difficulties in segmentation, we use the dense discrete points along the metal part edges as semantic keypoints for contour detection. Afterward, we exploit both keypoint information and CAD model to calculate the 6D pose of each object in view. A typical implementation of deep learning methods not only requires a large amount of training data, but also relies on intensive human labor for labeling the datasets. Therefore, we propose an approach to generate datasets and label them automatically. Despite not using any real-world photos for training, a series of experiments showed that the algorithm built on synthetic data perform well in the real environment.

Donghui Ding, Xin Jiang, Jiaping Wang et al.

Current blockchains are restricted by the low throughput. Aimed at this problem, we propose Txilm, a protocol that compresses the size of transaction presentation in each block to save the bandwidth of the network. In this protocol, a block carries short hashes of TXIDs instead of complete transactions. Combined with the sorted transactions based on TXIDs, Txilm realizes 80 times of data size reduction compared with the original blockchains. We also evaluate the probability of hash collisions, and provide methods of resolving such collisions. Finally, we design strategies to protect against potential attacks on Txilm.