Drug Target Interaction (DTI) prediction has witnessed promising performance boosts accompanied by advanced multimodal feature extraction. However, existing approaches suffer from two main difficulties. First, the complex protein structures cannot be well represented by current protein-sequence-based feature extractors. Second, the gap between protein and drug features increases the vulnerability of the obtained classifier thus degrading the prediction robustness. To address these issues, we propose a novel R-DTI method by exploring the second-order relevance in both protein structural feature extraction and DTI prediction phases. Specifically, we construct a pre-trained structural feature extractor that mines the atomic relevance of each amino acid. Then, an inter-feature structure-preserved Riemannian network is designed to expand the existing protein extraction patterns. To improve the prediction robustness, we also develop a Riemannian classifier that uses the second-order protein-drug relevance with a unified feature space. Extensive experimental results demonstrate the merits and superiority of our R-DTI against the state-of-the-art, achieving 1.4% and 1.9% higher AUC-ROC on the BindingDB and DrugBank datasets, respectively.

Model-based methods, which define pose parameters as 3D rotations of 24 joints relative to their parent joints, have proven highly effective in recovering 3D human meshes. Existing methods typically estimate these parameters by coupling features extracted from 2D images. However, the performance of these methods is always limited by spatial ambiguities due to dimensional inconsistencies and kinematic chain ambiguities caused by direct regression of all parameters. To overcome these limitations, we present a novel method that uses 3D joint-aware features within a GRU-based kinematic chain (AFGK) framework to obtain spatially accurate features. Specifically, we first use a self-attention mechanism to enable mutual interactions between features around keypoints, generating 3D joint-aware features, improving spatial accuracy of keypoints, and robustness to extreme poses. Then, a GRU-structured Kinematic chain update module (GRK) is introduced to decouple and update the 24 pose parameters. This module captures the kinematic relationships between keypoints and uses the features of parent joints to dynamically update the pose parameters of child joints, further reducing relative rotation errors. Extensive experimental results show that our method outperforms state-of-the-art competitors on standard benchmarks, significantly improving human reconstruction. The codes will be available on the project homepage: https://github.com/jingchenzhang/AFGK.

Recently, depth ambiguity reduction has made significant progress in monocular human mesh reconstruction. However, regression-based methods struggle to perceive minor depth deviations due to insufficiently exploited monocular cues, such as linear perspective, shadows, and pose priors of the given image. To tackle this limitation, we propose a Depth-aware Updated Feature Feedback (DAUFF) loop to iteratively exploit spatial information from rendered images. Specifically, the normalized depth and dense correspondence images rendered from the ground truth human meshes are adopted as auxiliary supervision for more reliable and adequate cues. Moreover, we also use dense correspondences for feature update to improve the adaptability to new predictions in the loop. Remarkably, dense correspondences can facilitate the rectification of parameters by providing a fine-grained perception of the current prediction. Extensive quantitative comparison on standard benchmarks shows that DAUFF achieves better-aligned reconstruction results than existing approaches. The dataset and codes will be available on the project homepage: https://github.com/github1838/DAUFF.