Dataset Condensation via Expert Subspace Projection.

The rapid growth in dataset sizes in modern deep learning has significantly increased data storage costs. Furthermore, the training and time costs for deep neural networks are generally proportional to the dataset size. Therefore, reducing the dataset size while maintaining model performance is an urgent research problem that needs to be addressed. Dataset condensation is a technique that aims to distill the original dataset into a much smaller synthetic dataset while maintaining downstream training performance on any agnostic neural network. Previous work has demonstrated that matching the training trajectory between the synthetic dataset and the original dataset is more effective than matching the instantaneous gradient, as it incorporates long-range information. Despite the effectiveness of trajectory matching, it suffers from complex gradient unrolling across iterations, which leads to significant memory and computation overhead. To address this issue, this paper proposes a novel approach called Expert Subspace Projection (ESP), which leverages long-range information while avoiding gradient unrolling. Instead of strictly enforcing the synthetic dataset's training trajectory to mimic that of the real dataset, ESP only constrains it to lie within the subspace spanned by the training trajectory of the real dataset. The memory-saving advantage offered by our method facilitates unbiased training on the complete set of synthetic images and seamless integration with other dataset condensation techniques. Through extensive experiments, we have demonstrated the effectiveness of our approach. Our method outperforms the trajectory matching method on CIFAR10 by 16.7% in the setting of 1 Image/Class, surpassing the previous state-of-the-art method by 3.2%.

A Weight-Adaptive Laplacian Embedding for Graph-Based Clustering.

Graph-based clustering methods perform clustering on a fixed input data graph. Thus such clustering results are sensitive to the particular graph construction. If this initial construction is of low quality, the resulting clustering may also be of low quality. We address this drawback by allowing the data graph itself to be adaptively adjusted in the clustering procedure. In particular, our proposed weight adaptive Laplacian (WAL) method learns a new data similarity matrix that can adaptively adjust the initial graph according to the similarity weight in the input data graph. We develop three versions of these methods based on the L2-norm, fuzzy entropy regularizer, and another exponential-based weight strategy, that yield three new graph-based clustering objectives. We derive optimization algorithms to solve these objectives. Experimental results on synthetic data sets and real-world benchmark data sets exhibit the effectiveness of these new graph-based clustering methods.

Melatonin reverses H2 O2 -induced premature senescence in mesenchymal stem cells via the SIRT1-dependent pathway.

Mesenchymal stem cells (MSCs) represent an attractive source for stem cell-based regenerative therapy, but they are vulnerable to oxidative stress-induced premature senescence in pathological conditions. We previously reported antioxidant and antiarthritic effects of melatonin on MSCs against proinflammatory cytokines. In this study, we hypothesized that melatonin could protect MSCs from premature senescence induced by hydrogen peroxide (H2 O2 ) via the silent information regulator type 1 (SIRT1)-dependent pathway. In response to H2 O2 at a sublethal concentration of 200 µm, human bone marrow-derived MSCs (BM-MSCs) underwent growth arrest and cellular senescence. Treatment with melatonin before H2 O2 exposure cannot significantly prevent premature senescence; however, treatment with melatonin subsequent to H2 O2 exposure successfully reversed the senescent phenotypes of BM-MSCs in a dose-dependent manner. This result was made evident by improved cell proliferation, decreased senescence-associated ß-galactosidase activity, and the improved entry of proliferating cells into the S phase. In addition, treatment with 100 µm melatonin restored the osteogenic differentiation potential of BM-MSCs that was inhibited by H2 O2 -induced premature senescence. We also found that melatonin attenuated the H2 O2 -stimulated phosphorylation of p38 mitogen-activated protein kinase, decreased expression of the senescence-associated protein p16(INK) (4α) , and increased SIRT1. Further molecular experiments revealed that luzindole, a nonselective antagonist of melatonin receptors, blocked melatonin-mediated antisenescence effects. Inhibition of SIRT1 by sirtinol counteracted the protective effects of melatonin, suggesting that melatonin reversed the senescence in cells through the SIRT1-dependent pathway. Together, these findings lay new ground for understanding oxidative stress-induced premature senescence and open perspectives for therapeutic applications of melatonin in stem cell-based regenerative medicine.

A Differential Protein Study on Bronchoalveolar Lavage Fluid at Different Stages of Silicosis.

OBJECTIVES: In this study, by comparing the difference in protein expression in bronchoalveolar lavage fluid between silicosis patients in different stages and healthy controls, the pathogenesis of pneumoconiosis was discussed, and a new idea for the prevention and treatment of pneumoconiosis was provided. METHODS: The lung lavage fluid was pretreated by 10 K ultrafiltration tube, Agilent 1100 conventional liquid phase separation, strong cation exchange column (SCX) HPLC pre-separation, and C18 reverse phase chromatography desalting purification, and protein was labeled with isotope. GO, KEGG pathway, and PPI analysis of differential proteins were conducted by bioinformatics, and protein types and corresponding signal pathways were obtained. RESULTS: Thermo Q-Exactive mass spectrometry identified 943 proteins. T-test analysis was used to evaluate the different significance of the results, and the different protein of each group was obtained by screening with the Ratio≥1.2 or Ratio≤0.83 and P<0.05. We found that there are 16 kinds of protein throughout the process of silicosis. There are different expressions of protein in stages Ⅲ/control, stages Ⅱ/control, stage Ⅰ/control, stages Ⅲ/ stages Ⅱ, stages Ⅲ/ stage Ⅰ and stages Ⅱ/ stage Ⅰ groups. The results of ontology enrichment analysis of total differential protein genes show that KEGG pathway enrichment analysis of differential protein suggested that there were nine pathways related to silicosis. CONCLUSION: The main biological changes in the early stage of silicosis are glycolysis or gluconeogenesis, autoimmunity, carbon metabolism, phagocytosis, etc., and microfibril-associated glycoprotein 4 may be involved in the early stage of silicosis. The main biological changes in the late stage of silicosis are autoimmunity, intercellular adhesion, etc. Calcium hippocampus binding protein may participate in the biological changes in the late stage of silicosis. It provides a new idea to understand the pathogenesis of silicosis and also raises new questions for follow-up research.

Melatonin mediates protective effects on inflammatory response induced by interleukin-1 beta in human mesenchymal stem cells.

Joint diseases like osteoarthritis usually are accompanied with inflammatory processes, in which pro-inflammatory cytokines mediate the generation of intracellular reactive oxygen species (ROS) and compromise survival of subchondral osteoblasts. Melatonin is capable of manipulating bone formation and osteogenic differentiation of mesenchymal stem cells (MSCs). The aim of this work was to investigate the anti-inflammatory effect of melatonin on MSC proliferation and osteogenic differentiation in the absence or presence of interleukin-1 beta (IL-1ß), which was used to induce inflammation. Our data showed that melatonin improved cell viability and reduced ROS generation in MSCs in a dose-dependent manner. When exposed to 10 ng/mL IL-1ß, various concentrations of melatonin resulted in significant reduction of ROS by 34.9% averagely. Luzindole as a melatonin receptor antagonist reversed the anti-oxidant effect of melatonin in MSCs with co-exposure to IL-1ß. Real-time RT-PCR data suggested that melatonin treatment up-regulated the expression of CuZnSOD and MnSOD, while down-regulated the expression of Bax. To investigate the effect of melatonin on osteogenesis, MSCs were cultured in osteogenic differentiation medium supplemented with IL-1ß, melatonin, or luzindole. After exposed to IL-1ß for 21 days, 1 µm melatonin treatment significantly increased the levels of type I collagen, ALP, and osteocalcin, and 100 µm melatonin treatment yielded the highest level of osteopontin. Our study demonstrated that melatonin maintained MSC survival and promoted osteogenic differentiation in inflammatory environment induced by IL-1ß, suggesting melatonin treatment could be a promising method for bone regenerative engineering in future studies.

Model Behavior Preserving for Class-Incremental Learning.

Deep models have shown to be vulnerable to catastrophic forgetting, a phenomenon that the recognition performance on old data degrades when a pre-trained model is fine-tuned on new data. Knowledge distillation (KD) is a popular incremental approach to alleviate catastrophic forgetting. However, it usually fixes the absolute values of neural responses for isolated historical instances, without considering the intrinsic structure of the responses by a convolutional neural network (CNN) model. To overcome this limitation, we recognize the importance of the global property of the whole instance set and treat it as a behavior characteristic of a CNN model relevant to model incremental learning. On this basis: 1) we design an instance neighborhood-preserving (INP) loss to maintain the order of pair-wise instance similarities of the old model in the feature space; 2) we devise a label priority-preserving (LPP) loss to preserve the label ranking lists within instance-wise label probability vectors in the output space; and 3) we introduce an efficient derivable ranking algorithm for calculating the two loss functions. Extensive experiments conducted on CIFAR100 and ImageNet show that our approach achieves the state-of-the-art performance.

Semi-Supervised Crowd Counting via Multiple Representation Learning.

There has been a growing interest in counting crowds through computer vision and machine learning techniques in recent years. Despite that significant progress has been made, most existing methods heavily rely on fully-supervised learning and require a lot of labeled data. To alleviate the reliance, we focus on the semi-supervised learning paradigm. Usually, crowd counting is converted to a density estimation problem. The model is trained to predict a density map and obtains the total count by accumulating densities over all the locations. In particular, we find that there could be multiple density map representations for a given image in a way that they differ in probability distribution forms but reach a consensus on their total counts. Therefore, we propose multiple representation learning to train several models. Each model focuses on a specific density representation and utilizes the count consistency between models to supervise unlabeled data. To bypass the explicit density regression problem, which makes a strong parametric assumption on the underlying density distribution, we propose an implicit density representation method based on the kernel mean embedding. Extensive experiments demonstrate that our approach outperforms state-of-the-art semi-supervised methods significantly.

Brain Cognition-Inspired Dual-Pathway CNN Architecture for Image Classification.

Inspired by the global-local information processing mechanism in the human visual system, we propose a novel convolutional neural network (CNN) architecture named cognition-inspired network (CogNet) that consists of a global pathway, a local pathway, and a top-down modulator. We first use a common CNN block to form the local pathway that aims to extract fine local features of the input image. Then, we use a transformer encoder to form the global pathway to capture global structural and contextual information among local parts in the input image. Finally, we construct the learnable top-down modulator where fine local features of the local pathway are modulated by global representations of the global pathway. For ease of use, we encapsulate the dual-pathway computation and modulation process into a building block, called the global-local block (GL block), and a CogNet of any depth can be constructed by stacking a necessary number of GL blocks one after another. Extensive experimental evaluations have revealed that the proposed CogNets have achieved the state-of-the-art performance accuracies on all the six benchmark datasets and are very effective for overcoming the "texture bias" and the "semantic confusion" problems faced by many CNN models.

Pulmonary decellularized extracellular matrix (dECM) modified polyethylene terephthalate three-dimensional cell carriers regulate the proliferation and paracrine activity of mesenchymal stem cells.

Introduction: Mesenchymal stem cells (MSCs) possess a high degree of self-renewal capacity and in vitro multi-lineage differentiation potential. Decellularized materials have garnered considerable attention due to their elevated biocompatibility, reduced immunogenicity, excellent biodegradability, and the ability to partially mimic the in vivo microenvironment conducive to cell growth. To address the issue of mesenchymal stem cells losing their stem cell characteristics during two-dimensional (2D) cultivation, this study established three-dimensional cell carriers modified with lung decellularized extracellular matrix and assessed its impact on the life activities of mesenchymal stem cells. Methods: This study employed PET as a substrate material, grafting with polydopamine (PDA), and constructing a decellularized extracellular matrix (dECM) coating on its surface, thus creating the PET/PDA/dECM three-dimensional (3D) composite carrier. Subsequently, material characterization of the cellular carriers was conducted, followed by co-culturing with human umbilical cord mesenchymal stem cells in vitro, aiming to investigate the material's impact on the proliferation and paracrine activity of mesenchymal stem cells. Results and Discussion: Material characterization demonstrated successful grafting of PDA and dECM materials, and it had complete hydrophilicity, high porosity, and excellent mechanical properties. The material was rich in various ECM proteins (collagen I, collagen IV , laminin, fibronectin, elastin), indicating good biocompatibility. In long-term in vitro cultivation (14 days) experiments, the PET/PDA/dECM three-dimensional composite carrier significantly enhanced adhesion and proliferation of human umbilical cord-derived mesenchymal stem cells (HUCMSCs), with a proliferation rate 1.9 times higher than that of cells cultured on tissue culture polystyrene (TCPS) at day 14. Furthermore, it effectively maintained the stem cell characteristics, expressing specific antigens for HUCMSCs. Through qPCR, Western blot, and ELISA experiments, the composite carrier markedly promoted the expression and secretion of key cell factors in HUCMSCs. These results demonstrate that the PET/PDA/dECM composite carrier holds great potential for scaling up MSCs' long-term in vitro cultivation and the production of paracrine factors.

Transductive Semisupervised Deep Hashing.

Deep hashing methods have shown their superiority to traditional ones. However, they usually require a large amount of labeled training data for achieving high retrieval accuracies. We propose a novel transductive semisupervised deep hashing (TSSDH) method which is effective to train deep convolutional neural network (DCNN) models with both labeled and unlabeled training samples. TSSDH method consists of the following four main ingredients. First, we extend the traditional transductive learning (TL) principle to make it applicable to DCNN-based deep hashing. Second, we introduce confidence levels for unlabeled samples to reduce adverse effects from uncertain samples. Third, we employ a Gaussian likelihood loss for hash code learning to sufficiently penalize large Hamming distances for similar sample pairs. Fourth, we design the large-margin feature (LMF) regularization to make the learned features satisfy that the distances of similar sample pairs are minimized and the distances of dissimilar sample pairs are larger than a predefined margin. Comprehensive experiments show that the TSSDH method can produce superior image retrieval accuracies compared to the representative semisupervised deep hashing methods under the same number of labeled training samples.

Identity-Quantity Harmonic Multi-Object Tracking.

The data association problem of multi-object tracking (MOT) aims to assign IDentity (ID) labels to detections and infer a complete trajectory for each target. Most existing methods assume that each detection corresponds to a unique target and thus cannot handle situations when multiple targets occur in a single detection due to detection failure in crowded scenes. To relax this strong assumption for practical applications, we formulate the MOT as a Maximizing An Identity-Quantity Posterior (MAIQP) problem on the basis of associating each detection with an identity and a quantity characteristic and then provide solutions to tackle two key problems arising. Firstly, a local target quantification module is introduced to count the number of targets within one detection. Secondly, we propose an identity-quantity harmony mechanism to reconcile the two characteristics. On this basis, we develop a novel Identity-Quantity HArmonic Tracking (IQHAT) framework that allows assigning multiple ID labels to detections containing several targets. Through extensive experimental evaluations on five benchmark datasets, we demonstrate the superiority of the proposed method.

Deep Class-Incremental Learning From Decentralized Data.

In this article, we focus on a new and challenging decentralized machine learning paradigm in which there are continuous inflows of data to be addressed and the data are stored in multiple repositories. We initiate the study of data-decentralized class-incremental learning (DCIL) by making the following contributions. First, we formulate the DCIL problem and develop the experimental protocol. Second, we introduce a paradigm to create a basic decentralized counterpart of typical (centralized) CIL approaches, and as a result, establish a benchmark for the DCIL study. Third, we further propose a decentralized composite knowledge incremental distillation (DCID) framework to transfer knowledge from historical models and multiple local sites to the general model continually. DCID consists of three main components, namely, local CIL, collaborated knowledge distillation (KD) among local models, and aggregated KD from local models to the general one. We comprehensively investigate our DCID framework by using a different implementation of the three components. Extensive experimental results demonstrate the effectiveness of our DCID framework. The source code of the baseline methods and the proposed DCIL is available at https://github.com/Vision-Intelligence-and-Robots-Group/DCIL.

A novel shell-structure cell microcarrier (SSCM) for cell transplantation and bone regeneration medicine.

PURPOSE: The present study aims to develop a novel open and hollow shell-structure cell microcarrier (SSCM) to improve the anchorage-dependent cell (ADC) loading efficiency, increase the space for cell proliferation and tissue regeneration, and better propel its therapeutic effects. METHODS: Gelatin particles were prepared with oil/water/oil (o/w/o) technique and modified by an adjustable surface crosslinking technique and subsequent release of uncrosslinked material. Optical microscopy and scanning electron microscopy (SEM) were utilized to observe the morphologies of the microcarriers. Cell loading tests were performed to evaluate the biocompatibilities and effect on osteogenesis of SSCM. RESULTS: SSCMs were successfully fabricated via the surface technique. The shell-structure could allow the cell to attach and grow on both outer and inner surface of sphere and provide adequate space for cell proliferation and extracellular matrix (ECM) secretion. The cell loading rate, proliferation rate and osteogenesis-related gene expressions on the SSCMs were higher than those on the spherical gelatin microcarriers. CONCLUSIONS: The outstanding performance of injectable SSCMs endowed with favorable micro-structure, desirable cytocompatibility and enhanced cell affinity makes them as a good choice as cell delivery vehicle for transplanting therapeutic cells towards the scope of tissue regeneration.

Analogy-Detail Networks for Object Recognition.

The human visual system can recognize object categories accurately and efficiently and is robust to complex textures and noises. To mimic the analogy-detail dual-pathway human visual cognitive mechanism revealed in recent cognitive science studies, in this article, we propose a novel convolutional neural network (CNN) architecture named analogy-detail networks (ADNets) for accurate object recognition. ADNets disentangle the visual information and process them separately using two pathways: the analogy pathway extracts coarse and global features representing the gist (i.e., shape and topology) of the object, while the detail pathway extracts fine and local features representing the details (i.e., texture and edges) for determining object categories. We modularize the architecture and encapsulate the two pathways into the analogy-detail block as the CNN building block to construct ADNets. For implementation, we propose a general principle that transmutes typical CNN structures into the ADNet architecture and applies the transmutation on representative baseline CNNs. Extensive experiments on CIFAR10, CIFAR100, street view house numbers, and ImageNet data sets demonstrate that ADNets significantly reduce the test error rates of the baseline CNNs by up to 5.76% and outperform other state-of-the-art architectures. Comprehensive analysis and visualizations further demonstrate that ADNets are interpretable and have a better shape-texture tradeoff for recognizing the objects with complex textures.

Fusion of Multiple Person Re-id Methods With Model and Data-Aware Abilities.

Person re-identification (person re-id) has attracted rapidly increasing attention in computer vision and pattern recognition research community in recent years. With the goal of providing match ranking results between each query person image and the gallery ones, the person re-id technique has been widely explored and a large number of person re-id methods have been developed. As these algorithms leverage different kinds of prior assumptions, image features, distance matching functions, et al., each of them has its own strengths and weaknesses. Inspired by these facts, this paper proposes a novel person re-id method based on the idea of inferring superior fusion results from a variety of previous base person re-id algorithms using different methodologies or features. To achieve this goal, we propose a novel framework which mainly consists of two steps: 1) a number of existing person re-id methods are implemented, and the ranking results are obtained in the test datasets. and 2) the robust fusion strategy is applied to obtain better re-ranked matching results by simultaneously considering the recognition abilities of various base re-id methods and the difficulties of different gallery person images to be correctly recognized under the generative model of labels, abilities, and difficulties framework. Comprehensive experiments show the effectiveness of our proposed method, and we have received state-of-the-art results on recent popular person re-id datasets.

SoftCuts: a soft edge smoothness prior for color image super-resolution.

Designing effective image priors is of great interest to image super-resolution (SR), which is a severely under-determined problem. An edge smoothness prior is favored since it is able to suppress the jagged edge artifact effectively. However, for soft image edges with gradual intensity transitions, it is generally difficult to obtain analytical forms for evaluating their smoothness. This paper characterizes soft edge smoothness based on a novel SoftCuts metric by generalizing the Geocuts method . The proposed soft edge smoothness measure can approximate the average length of all level lines in an intensity image. Thus, the total length of all level lines can be minimized effectively by integrating this new form of prior. In addition, this paper presents a novel combination of this soft edge smoothness prior and the alpha matting technique for color image SR, by adaptively normalizing image edges according to their alpha-channel description. This leads to the adaptive SoftCuts algorithm, which represents a unified treatment of edges with different contrasts and scales. Experimental results are presented which demonstrate the effectiveness of the proposed method.

Fine-Grained Image Classification Using Modified DCNNs Trained by Cascaded Softmax and Generalized Large-Margin Losses.

We develop a fine-grained image classifier using a general deep convolutional neural network (DCNN). We improve the fine-grained image classification accuracy of a DCNN model from the following two aspects. First, to better model the h -level hierarchical label structure of the fine-grained image classes contained in the given training data set, we introduce h fully connected (fc) layers to replace the top fc layer of a given DCNN model and train them with the cascaded softmax loss. Second, we propose a novel loss function, namely, generalized large-margin (GLM) loss, to make the given DCNN model explicitly explore the hierarchical label structure and the similarity regularities of the fine-grained image classes. The GLM loss explicitly not only reduces between-class similarity and within-class variance of the learned features by DCNN models but also makes the subclasses belonging to the same coarse class be more similar to each other than those belonging to different coarse classes in the feature space. Moreover, the proposed fine-grained image classification framework is independent and can be applied to any DCNN structures. Comprehensive experimental evaluations of several general DCNN models (AlexNet, GoogLeNet, and VGG) using three benchmark data sets (Stanford car, fine-grained visual classification-aircraft, and CUB-200-2011) for the fine-grained image classification task demonstrate the effectiveness of our method.

Discriminative Feature Learning with Foreground Attention for Person Re-identification.

The performance of person re-identification (Re-ID) has been seriously effected by the large cross-view appearance variations caused by mutual occlusions and background clutters. Hence learning a feature representation that can adaptively emphasize the foreground persons becomes very critical to solve the person Re-ID problem. In this paper, we propose a simple yet effective foreground attentive neural network (FANN) to learn a discriminative feature representation for person Re-ID, which can adaptively enhance the positive side of foreground and weaken the negative side of background. Specifically, a novel foreground attentive subnetwork is designed to drive the network's attention, in which a decoder network is used to reconstruct the binary mask by using a novel local regression loss function, and an encoder network is regularized by the decoder network to focus its attention on the foreground persons. The resulting feature maps of encoder network are further fed into the body part subnetwork and feature fusion subnetwork to learn discriminative features. Besides, a novel symmetric triplet loss function is introduced to supervise feature learning, in which the intra-class distance is minimized and the inter-class distance is maximized in each triplet unit, simultaneously. Training our FANN in a multi-task learning framework, a discriminative feature representation can be learned to find out the matched reference to each probe among various candidates in the gallery. Extensive experimental results on several public benchmark datasets are evaluated, which have shown clear improvements of our method over the state-of-the-art approaches.

Development of CaCO3 microsphere-based composite hydrogel for dual delivery of growth factor and Ca to enhance bone regeneration.

Injectable scaffolds have attracted much attention because of their minimum surgical invasiveness. However, limited osteogenic induction property and low mechanical properties hampered their application in bone tissue engineering. CaCO3 microspheres, which possess osteoinductivity, rough surfaces and specific binding sites for BMP-2, were first fabricated; after BMP-2 uploading, microspheres were further entrapped in fibrin-glue hydrogel. CaCO3 microspheres were co-functionalized with casein and heparin. To obtain a high encapsulation of heparin and thus BMP-2 uploading, along with controlled release and simultaneous maintenance of the presence of vaterite which had osteogenic induction property, fabrication parameters were optimized and microspheres were characterized using XRD, FITR and SEM. The formed CaCO3 had a microsphere morphology of ∼1 µm. Both vaterite and calcite phases were present and the relative amount of calcite phase increased with the amount of heparin. Sample 25 mM_4-1Hep with the highest loading amount of heparin was selected as carrier for BMP-2 and BMP-2 loaded CaCO3 microspheres were further entrapped in fibrin-glue hydrogel (FC-B). For the as-prepared composite hydrogel, mechanical properties were characterized and the presence of CaCO3 significantly elevated the tensile strength; controlled release of BMP-2 was sustained until day 21. Based on ALP activity, alizarin red staining and RT-PCR, in vitro osteogenic differentiation of bone marrow mesenchymal stem cells (BMSCs) was found to be significantly enhanced under induction of FC-B. Rabbit tibia bone defect model was applied to evaluate its in vivo performance. After implantation for 4 weeks, presence of composite hydrogel was observed in defects. After 8 weeks, bone defects of FC-B group were nearly completely healed. Using the fact that autologous scaffolds can be derived based on fibrin-glue hydrogel, the well-designed BMP-2 loaded fibrin-glue composite hydrogel demonstrated good potential in bone tissue engineering.

Enzymatic synthesis of PEG-poly(amine-co-thioether esters) as highly efficient pH and ROS dual-responsive nanocarriers for anticancer drug delivery.

Novel multifunctional drug nanocarriers have been successfully fabricated from a new type of enzymatically synthesized, biodegradable block copolymer, PEG-poly(ω-pentadecalactone-co-N-methyldiethyleneamine-co-3,3'-thiodipropionate) (PEG-PPMT), which was responsive to tumor-relevant acidic pH (5.0-6.5) and intracellular reactive oxygen species (ROS) of tumor cells. The PEG-PPMT copolymers could self-assemble to form nano-scaled particles in aqueous solutions, which are stable in physiological solutions, but swell substantially upon reducing the pH from 7.4 to 5.0 and/or in the presence of ROS on account of the protonation of the tertiary amino groups and oxidation of the thioether groups, causing a hydrophobic to hydrophilic transition in the nanoparticle cores. Consistently, docetaxel (DTX) encapsulated in PEG-PPMT nanoparticles can be triggered in a synergistic manner by acidic pH and a high-ROS environment in tumor cells to release the hydrophobic drug at accelerated rates for efficient tumor growth inhibition. In particular, DTX encapsulated in PEG-PPMT-11% PDL and PEG-PPMT-28% PDL nanoparticles exhibit extraordinarily enhanced potency (95% and 93% tumor-inhibiting efficiency, respectively) in inhibiting the growth of ROS-rich CT-26 tumors xenografted in mice. Importantly, biosafety analyses show minimal toxicity of DTX-loaded PEG-PPMT nanoparticles toward normal organs including liver and kidneys during the in vivo antitumor treatments. These results demonstrate that the PEG-PPMT nanoparticles are promising pH and ROS dual-responsive multifunctional nanocarriers for tumor site specific, controlled release of anticancer drugs to treat ROS-rich tumors.