Spatiotemporal Decouple-and-Squeeze Contrastive Learning for Semisupervised Skeleton-Based Action Recognition.

Contrastive learning has been successfully leveraged to learn action representations for addressing the problem of semisupervised skeleton-based action recognition. However, most contrastive learning-based methods only contrast global features mixing spatiotemporal information, which confuses the spatial-and temporal-specific information reflecting different semantic at the frame level and joint level. Thus, we propose a novel spatiotemporal decouple-and-squeeze contrastive learning (SDS-CL) framework to comprehensively learn more abundant representations of skeleton-based actions by jointly contrasting spatial-squeezing features, temporal-squeezing features, and global features. In SDS-CL, we design a new spatiotemporal-decoupling intra-inter attention (SIIA) mechanism to obtain the spatiotemporal-decoupling attentive features for capturing spatiotemporal specific information by calculating spatial-and temporal-decoupling intra-attention maps among joint/motion features, as well as spatial-and temporal-decoupling inter-attention maps between joint and motion features. Moreover, we present a new spatial-squeezing temporal-contrasting loss (STL), a new temporal-squeezing spatial-contrasting loss (TSL), and the global-contrasting loss (GL) to contrast the spatial-squeezing joint and motion features at the frame level, temporal-squeezing joint and motion features at the joint level, as well as global joint and motion features at the skeleton level. Extensive experimental results on four public datasets show that the proposed SDS-CL achieves performance gains compared with other competitive methods.

Multi-Granularity Anchor-Contrastive Representation Learning for Semi-Supervised Skeleton-Based Action Recognition.

In the semi-supervised skeleton-based action recognition task, obtaining more discriminative information from both labeled and unlabeled data is a challenging problem. As the current mainstream approach, contrastive learning can learn more representations of augmented data, which can be considered as the pretext task of action recognition. However, such a method still confronts three main limitations: 1) It usually learns global-granularity features that cannot well reflect the local motion information. 2) The positive/negative pairs are usually pre-defined, some of which are ambiguous. 3) It generally measures the distance between positive/negative pairs only within the same granularity, which neglects the contrasting between the cross-granularity positive and negative pairs. Toward these limitations, we propose a novel Multi-granularity Anchor-Contrastive representation Learning (dubbed as MAC-Learning) to learn multi-granularity representations by conducting inter- and intra-granularity contrastive pretext tasks on the learnable and structural-link skeletons among three types of granularities covering local, context, and global views. To avoid the disturbance of ambiguous pairs from noisy and outlier samples, we design a more reliable Multi-granularity Anchor-Contrastive Loss (dubbed as MAC-Loss) that measures the agreement/disagreement between high-confidence soft-positive/negative pairs based on the anchor graph instead of the hard-positive/negative pairs in the conventional contrastive loss. Extensive experiments on both NTU RGB+D and Northwestern-UCLA datasets show that the proposed MAC-Learning outperforms existing competitive methods in semi-supervised skeleton-based action recognition tasks.

X-Invariant Contrastive Augmentation and Representation Learning for Semi-Supervised Skeleton-Based Action Recognition.

Semi-supervised skeleton-based action recognition is a challenging problem due to insufficient labeled data. For addressing this problem, some representative methods leverage contrastive learning to obtain more features from the pre-augmented skeleton actions. Such methods usually adopt a two-stage way: first randomly augment samples, and then learn their representations via contrastive learning. Since skeleton samples have already been randomly augmented, the representation ability of the subsequent contrastive learning is limited due to the inconsistency between the augmentations and representations. Thus, we propose a novel X-invariant Contrastive Augmentation and Representation learning (X-CAR) framework to thoroughly obtain rotate-shear-scale (X for short) invariant features by learning augmentations and representations of skeleton sequences in a one-stage way. In X-CAR, a new Adaptive-combination Augmentation (AA) mechanism is designed to rotate, shear, and scale the skeletons by learnable controlling factors in an adaptive way rather than a random way. Here, such controlling factors are also learned in the whole contrastive learning process, which can facilitate the consistency between the learned augmentations and representations of skeleton sequences. In addition, we relax the pre-definition of positive and negative samples to avoid the confusing allocation of ambiguous samples, and present a new Pull-Push Contrastive Loss (PPCL) to pull the augmenting skeleton close to the original skeleton, while push far away from the other skeletons. Experimental results on both NTU RGB+D and North-Western UCLA datasets show that the proposed X-CAR achieves better accuracy compared with other competitive methods in the semi-supervised scenario.

His18 promotes reactive oxidative stress production in copper-ion mediated human islet amyloid polypeptide aggregation.

Copper ions play a critical role in human islet amyloid polypeptide (hIAPP) aggregation, which has been found in more than 90% of patients with type-2 diabetes (T2D). The role of Cu(ii) in the cell cytotoxicity with hIAPP has been explored in two aspects: inhibiting the formation of fibrillar structures and stimulating the generation of reactive oxygen species (ROS). In this work, we carried out spectroscopic studies of Cu(ii) interacting with several hIAPP fragments and their variants as well. Electron paramagnetic resonance (EPR) measurements and Amplex Red analysis showed that the amount of H2O2 generated in hIAPP(11-28) solution co-incubated with Cu(ii) was remarkably more than hIAPP(1-11) and hIAPP(28-37). Furthermore, the H2O2 level was seriously reduced when His18 of hIAPP(11-28) was replaced by Arg(R) or Ser(S), indicating that His18 is the key residue of Cu(ii) binding to hIAPP(11-28) to promote H2O2 generation. This is likely because the donation of electrons from the peptide to Cu(ii) ions would result in the formation of the redox-active complexes, which could stimulate the formation of H2O2. Overall, this study provides further insight into the molecular mechanism of Cu(ii) induced ROS generation.

Chemical imaging of protein hydrogels undergoing alkaline dissolution by CARS microscopy.

Hydrogels swell, shrink and degrade depending on the solution they are in contact which, strongly affecting their performance. The minimum information needed to validate many published simulations would be the spatial quantification of the solute material with time. In this study we develop a simple methodology to quantify the protein content in heat induced protein hydrogels using a commercial Coherent anti-Stokes Raman Spectroscopy (CARS) microscope. The system is used to quantify the whey protein isolate (WPI) concentration in hydrogels undergoing dissolution at alkaline pH. Quantitative measurements were performed in hydrogels up to depths of ∼600 µm, with an average accuracy of ∼1 wt%. Results show that the protein concentration within the swollen layer is constant with time, confirming the existence of steady state conditions during dissolution. The methodology presented can easily be implemented to other biopolymer hydrogels and foods.

Ferritin nanocages to encapsulate and deliver photosensitizers for efficient photodynamic therapy against cancer.

Photodynamic therapy is an emerging treatment modality that is under intensive preclinical and clinical investigations for many types of disease including cancer. Despite the promise, there is a lack of a reliable drug delivery vehicle that can transport photosensitizers (PSs) to tumors in a site-specific manner. Previous efforts have been focused on polymer- or liposome-based nanocarriers, which are usually associated with a suboptimal PS loading rate and a large particle size. We report herein that a RGD4C-modified ferritin (RFRT), a protein-based nanoparticle, can serve as a safe and efficient PS vehicle. Zinc hexadecafluorophthalocyanine (ZnF16Pc), a potent PS with a high (1)O2 quantum yield but poor water solubility, can be encapsulated into RFRTs with a loading rate as high as ~60 wt % (i.e., 1.5 mg of ZnF16Pc can be loaded on 1 mg of RFRTs), which far exceeds those reported previously. Despite the high loading, the ZnF16Pc-loaded RFRTs (P-RFRTs) show an overall particle size of 18.6 ± 2.6 nm, which is significantly smaller than other PS-nanocarrier conjugates. When tested on U87MG subcutaneous tumor models, P-RFRTs showed a high tumor accumulation rate (tumor-to-normal tissue ratio of 26.82 ± 4.07 at 24 h), a good tumor inhibition rate (83.64% on day 12), as well as minimal toxicity to the skin and other major organs. This technology can be extended to deliver other metal-containing PSs and holds great clinical translation potential.