Search

This work was developed in part under a contract from the US Department of Defense., Asian populations of gypsy moths, Lymantria dispar (L.) (Lepidoptera: Lymantriidae), remain poorly characterized – indeed, they are not presently accorded any formal taxonomic status within the broader species. Their ecology is similarly largely uncharacterized in the literature, except by assumption that it will resemble that of European populations. We developed ecological niche models specific to Asian populations of the species, which can in turn be used to identify a potential geographic distributional area for the species. We demonstrated statistically significant predictivity of distributional patterns within the East Asian range of these populations; projecting the Asian ecological niche model to Europe, correspondence with European distributions was generally good, although some differences may exist; projecting the ecological niche model globally, we characterized a likely potential invasive distribution of this set of populations across the temperate zone of both Northern and Southern Hemisphere., Depto. de Genética, Fisiología y Microbiología, Fac. de Ciencias Biológicas, TRUE, pub

We introduce VASA, a framework for generating lifelike talking faces with appealing visual affective skills (VAS) given a single static image and a speech audio clip. Our premiere model, VASA-1, is capable of not only producing lip movements that are exquisitely synchronized with the audio, but also capturing a large spectrum of facial nuances and natural head motions that contribute to the perception of authenticity and liveliness. The core innovations include a holistic facial dynamics and head movement generation model that works in a face latent space, and the development of such an expressive and disentangled face latent space using videos. Through extensive experiments including evaluation on a set of new metrics, we show that our method significantly outperforms previous methods along various dimensions comprehensively. Our method not only delivers high video quality with realistic facial and head dynamics but also supports the online generation of 512x512 videos at up to 40 FPS with negligible starting latency. It paves the way for real-time engagements with lifelike avatars that emulate human conversational behaviors., Comment: Tech Report. Project webpage: https://www.microsoft.com/en-us/research/project/vasa-1

Tumor-associated adipocytes (TAAs) recruit monocytes and promote their differentiation into tumor-associated macrophages (TAMs) that support tumor development. Here, TAAs are engineered to promote the polarization of TAMs to the tumor suppressive M1 phenotype. Telratolimod, a toll-like receptor 7/8 agonist, is loaded into the lipid droplets of adipocytes to be released at the tumor site upon tumor cell-triggered lipolysis. Locally administered drug-loaded adipocytes increased tumor suppressive M1 macrophages in both primary and distant tumors and suppressed tumor growth in a melanoma model. Furthermore, drug-loaded adipocytes improved CD8+ T cell-mediated immune responses within the tumor microenvironment and favored dendritic cell maturation in the tumor draining lymph nodes.

Commanding a drone with a natural language is not only user-friendly but also opens the door for emerging language agents to control the drone. Emerging large language models (LLMs) provide a previously impossible opportunity to automatically translate a task description in a natural language to a program that can be executed by the drone. However, powerful LLMs and their vision counterparts are limited in three important ways. First, they are only available as cloud-based services. Sending images to the cloud raises privacy concerns. Second, they are expensive, costing proportionally to the request size. Finally, without expensive fine-tuning, existing LLMs are quite limited in their capability of writing a program for specialized systems like drones. In this paper, we present a system called TypeFly that tackles the above three problems using a combination of edge-based vision intelligence, novel programming language design, and prompt engineering. Instead of the familiar Python, TypeFly gets a cloud-based LLM service to write a program in a small, custom language called MiniSpec, based on task and scene descriptions in English. Such MiniSpec programs are not only succinct (and therefore efficient) but also able to consult the LLM during their execution using a special skill called query. Using a set of increasingly challenging drone tasks, we show that design choices made by TypeFly can reduce both the cost of LLM service and the task execution time by more than 2x. More importantly, query and prompt engineering techniques contributed by TypeFly significantly improve the chance of success of complex tasks.

We present Prompt Cache, an approach for accelerating inference for large language models (LLM) by reusing attention states across different LLM prompts. Many input prompts have overlapping text segments, such as system messages, prompt templates, and documents provided for context. Our key insight is that by precomputing and storing the attention states of these frequently occurring text segments on the inference server, we can efficiently reuse them when these segments appear in user prompts. Prompt Cache employs a schema to explicitly define such reusable text segments, called prompt modules. The schema ensures positional accuracy during attention state reuse and provides users with an interface to access cached states in their prompt. Using a prototype implementation, we evaluate Prompt Cache across several LLMs. We show that Prompt Cache significantly reduce latency in time-to-first-token, especially for longer prompts such as document-based question answering and recommendations. The improvements range from 8x for GPU-based inference to 60x for CPU-based inference, all while maintaining output accuracy and without the need for model parameter modifications., Comment: To appear at MLSys 2024

This paper presents a novel neural implicit radiance representation for free viewpoint relighting from a small set of unstructured photographs of an object lit by a moving point light source different from the view position. We express the shape as a signed distance function modeled by a multi layer perceptron. In contrast to prior relightable implicit neural representations, we do not disentangle the different reflectance components, but model both the local and global reflectance at each point by a second multi layer perceptron that, in addition, to density features, the current position, the normal (from the signed distace function), view direction, and light position, also takes shadow and highlight hints to aid the network in modeling the corresponding high frequency light transport effects. These hints are provided as a suggestion, and we leave it up to the network to decide how to incorporate these in the final relit result. We demonstrate and validate our neural implicit representation on synthetic and real scenes exhibiting a wide variety of shapes, material properties, and global illumination light transport., Comment: Accepted to SIGGRAPH 2023. Author's version. Project page: https://nrhints.github.io

Bus headway deviations are the major causes of bus bunching, which is detrimental to bus service reliability and leads to excessive waiting time for passengers. Generally, it is believed that bus bunching is an inevitable result once a service disturbance occurs, as long as the route is long enough. However, in this study, it goes as low as 5% of all trips from the empirical data, contradicting the expectation in theory. This study explores the underlying reasons for this contradiction based on empirical data from two representative bus routes in China. The results show that headway deviations on links and at stops mostly counteract with each other, constraining the occurrence of bus bunching. Under this counteraction effect, the path headway deviations remain within a stable range, normally bounded by the planned headway. It indicates that the bus system has a self-repair ability to resist headway disruptions, which explains why bus bunching happens at a considerably lower frequency in practice than in theory. We further discuss how the counteraction effect works. The insights gained from this study can shed light on preventing bus bunching and generating new ways to recover bus service regularity after a headway disruption. If we can adequately adjust the control methods to make the link and stop headway deviations counteract with each other, the bus system can automatically mitigate the headway deviations. © 2022 Elsevier Ltd

Despite the promise of immune checkpoint blockade (ICB) therapy against cancer, challenges associated with low objective response rates and severe systemic side effects still remain and limit its clinical applications. Here, we described a cold atmospheric plasma (CAP)-mediated ICB therapy integrated with microneedles (MN) for the transdermal delivery of ICB. We found that a hollow-structured MN (hMN) patch facilitates the transportation of CAP through the skin, causing tumor cell death. The release of tumor-associated antigens then promotes the maturation of dendritic cells in the tumor-draining lymph nodes, subsequently initiating T cell-mediated immune response. Anti-programmed death-ligand 1 antibody (aPDL1), an immune checkpoint inhibitor, released from the MN patch further augments the antitumor immunity. Our findings indicate that the proposed transdermal combined CAP and ICB therapy can inhibit the tumor growth of both primary tumors and distant tumors, prolonging the survival of tumor-bearing mice.

Disruptions in the metro system often result in chaos in the public transport system given their substantial modal share in major metropolitan areas. Metro system disruption can be managed both within the system and by substitute bus (SB) service external to the system. Managing disruption within the metro system typically involves timetable adjustment, rolling stock scheduling, and crew rescheduling, subject to the rail system design and availability of crossovers for turnovers. SB service or bus bridging, on the other hand, is one prominent way to mitigate the negative impact of a metro disruption. In the past decade, research on this topic has grown substantially. In this paper, by conducting a comprehensive search in the Web of Science publications, we identify and review about 70 papers. According to the three aspects of metro system disruption management, we classify the literature into three groups: (i) getting prepared for metro network disruption, (ii) managing disruption within the metro system, and (iii) managing metro disruption with SB service. After summarising the state-of-the-art on this topic, we conclude by identifying gaps and future research directions, primarily on substitute bus service, for this important development effort. © 2020 Informa UK Limited, trading as Taylor & Francis Group.

The VirtualCube system is a 3D video conference system that attempts to overcome some limitations of conventional technologies. The key ingredient is VirtualCube, an abstract representation of a real-world cubicle instrumented with RGBD cameras for capturing the 3D geometry and texture of a user. We design VirtualCube so that the task of data capturing is standardized and significantly simplified, and everything can be built using off-the-shelf hardware. We use VirtualCubes as the basic building blocks of a virtual conferencing environment, and we provide each VirtualCube user with a surrounding display showing life-size videos of remote participants. To achieve real-time rendering of remote participants, we develop the V-Cube View algorithm, which uses multi-view stereo for more accurate depth estimation and Lumi-Net rendering for better rendering quality. The VirtualCube system correctly preserves the mutual eye gaze between participants, allowing them to establish eye contact and be aware of who is visually paying attention to them. The system also allows a participant to have side discussions with remote participants as if they were in the same room. Finally, the system sheds lights on how to support the shared space of work items (e.g., documents and applications) and track the visual attention of participants to work items., Comment: Project page: https://www.microsoft.com/en-us/research/project/virtualcube

Disruptions in the metro system often result in chaos in the public transport system given their substantial modal share in major metropolitan areas. Metro system disruption can be managed both within the system and by substitute bus (SB) service external to the system. Managing disruption within the metro system typically involves timetable adjustment, rolling stock scheduling, and crew rescheduling, subject to the rail system design and availability of crossovers for turnovers. SB service or bus bridging, on the other hand, is one prominent way to mitigate the negative impact of a metro disruption. In the past decade, research on this topic has grown substantially. In this paper, by conducting a comprehensive search in the Web of Science publications, we identify and review about 70 papers. According to the three aspects of metro system disruption management, we classify the literature into three groups: (i) getting prepared for metro network disruption, (ii) managing disruption within the metro system, and (iii) managing metro disruption with SB service. After summarising the state-of-the-art on this topic, we conclude by identifying gaps and future research directions, primarily on substitute bus service, for this important development effort. © 2020 Informa UK Limited, trading as Taylor & Francis Group.

The extremal inequality approach plays a key role in network information theory problems. In this paper, we propose a novel monotone path construction in product probability space. The optimality of Gaussian distribution is then established by standard perturbation arguments. The proofs of Liu-Viswanath extremal and vector Generalization of Costa's entropy power inequality are illustrated into the unified framework. As applications, capacity region of the multiple-input multiple-output (MIMO) Gaussian broadcast channel and rate-distortion-equivocation function of the vector Gaussian secure source coding are revisited through our proposed extremal inequality approach., Comment: 22 pages

We present POD, a smartphone that flies, as a new way to achieve hands-free, eyes-up mobile computing. Unlike existing drone-carried user interfaces, POD features a smartphone-sized display and the computing and sensing power of a modern smartphone. We share our experience in building a prototype of POD, discuss the technical challenges facing it, and describe early results toward addressing them.

In China, the bus system is frequently-serviced, and vehicles often queue at stops due to the limited capacity of a berth. Passengers will show out a disorder when boarding on buses, which may affect the service time at stops. In this paper, the effect of disorder, in the form the number of queues, on bus dwell time is quantified, and our research results show that when boarding passengers are organized by a single queue, the service time per passenger is 1.724 s/pax; while when passengers are organized by two queues, the service time per passenger is about 1.433 s/pax, decreased by about 16.88%. However, when passengers are organized by more than two queues, the service time per passenger increases sharply to 1.903 s/pax. It suggests that the service efficiency can be improved when boarding passengers are organized by two queues but will worsen with the aggravation of disorder. © ASCE.

Dwell time is an essential component of bus trip travel time, and it is affected by quantities of factors. To calculate the dwell time at bus stops, we must know the time when doors open and close, however, sometimes the events of doors open and closed are not recorded in GPS data. To solve this problem, we assume that there is a stable proportional relationship between dwell time and trip time, and if it is true, the estimation of dwell time can be greatly simplified. A proportion model, DT = kTT, was established in this research, and it was tested by more than 20,000 trips of route No. 210 located in Shanghai, and the calibrated coefficient k is equal to 0.261, varying little, among different time-of-days, and different month, which prove that we can use this proportionality to estimate the dwell time of bus route once trip travel time is available. © ASCE.

Disruptions in the metro system often result in chaos in the public transport system given their substantial modal share in major metropolitan areas. Metro system disruption can be managed both within the system and by substitute bus (SB) service external to the system. Managing disruption within the metro system typically involves timetable adjustment, rolling stock scheduling, and crew rescheduling, subject to the rail system design and availability of crossovers for turnovers. SB service or bus bridging, on the other hand, is one prominent way to mitigate the negative impact of a metro disruption. In the past decade, research on this topic has grown substantially. In this paper, by conducting a comprehensive search in the Web of Science publications, we identify and review about 70 papers. According to the three aspects of metro system disruption management, we classify the literature into three groups: (i) getting prepared for metro network disruption, (ii) managing disruption within the metro system, and (iii) managing metro disruption with SB service. After summarising the state-of-the-art on this topic, we conclude by identifying gaps and future research directions, primarily on substitute bus service, for this important development effort. © 2020 Informa UK Limited, trading as Taylor & Francis Group.

In China, the bus system is frequently-serviced, and vehicles often queue at stops due to the limited capacity of a berth. Passengers will show out a disorder when boarding on buses, which may affect the service time at stops. In this paper, the effect of disorder, in the form the number of queues, on bus dwell time is quantified, and our research results show that when boarding passengers are organized by a single queue, the service time per passenger is 1.724 s/pax; while when passengers are organized by two queues, the service time per passenger is about 1.433 s/pax, decreased by about 16.88%. However, when passengers are organized by more than two queues, the service time per passenger increases sharply to 1.903 s/pax. It suggests that the service efficiency can be improved when boarding passengers are organized by two queues but will worsen with the aggravation of disorder. © ASCE.

Dwell time is an essential component of bus trip travel time, and it is affected by quantities of factors. To calculate the dwell time at bus stops, we must know the time when doors open and close, however, sometimes the events of doors open and closed are not recorded in GPS data. To solve this problem, we assume that there is a stable proportional relationship between dwell time and trip time, and if it is true, the estimation of dwell time can be greatly simplified. A proportion model, DT = kTT, was established in this research, and it was tested by more than 20,000 trips of route No. 210 located in Shanghai, and the calibrated coefficient k is equal to 0.261, varying little, among different time-of-days, and different month, which prove that we can use this proportionality to estimate the dwell time of bus route once trip travel time is available. © ASCE.

Dwell time is an essential component of bus trip travel time, and it is affected by quantities of factors. To calculate the dwell time at bus stops, we must know the time when doors open and close, however, sometimes the events of doors open and closed are not recorded in GPS data. To solve this problem, we assume that there is a stable proportional relationship between dwell time and trip time, and if it is true, the estimation of dwell time can be greatly simplified. A proportion model, DT = kTT, was established in this research, and it was tested by more than 20,000 trips of route No. 210 located in Shanghai, and the calibrated coefficient k is equal to 0.261, varying little, among different time-of-days, and different month, which prove that we can use this proportionality to estimate the dwell time of bus route once trip travel time is available. © ASCE.

In China, the bus system is frequently-serviced, and vehicles often queue at stops due to the limited capacity of a berth. Passengers will show out a disorder when boarding on buses, which may affect the service time at stops. In this paper, the effect of disorder, in the form the number of queues, on bus dwell time is quantified, and our research results show that when boarding passengers are organized by a single queue, the service time per passenger is 1.724 s/pax; while when passengers are organized by two queues, the service time per passenger is about 1.433 s/pax, decreased by about 16.88%. However, when passengers are organized by more than two queues, the service time per passenger increases sharply to 1.903 s/pax. It suggests that the service efficiency can be improved when boarding passengers are organized by two queues but will worsen with the aggravation of disorder. © ASCE.

We present a scheme for fast environment light estimation from the RGBD appearance of individual objects and their local image areas. Conventional inverse rendering is too computationally demanding for real-time applications, and the performance of purely learning-based techniques may be limited by the meager input data available from individual objects. To address these issues, we propose an approach that takes advantage of physical principles from inverse rendering to constrain the solution, while also utilizing neural networks to expedite the more computationally expensive portions of its processing, to increase robustness to noisy input data as well as to improve temporal and spatial stability. This results in a rendering-aware system that estimates the local illumination distribution at an object with high accuracy and in real time. With the estimated lighting, virtual objects can be rendered in AR scenarios with shading that is consistent to the real scene, leading to improved realism., Comment: ECCV 2020

In this paper, the rate-distortion theory of the Gray-Wyner lossy source coding system is investigated. For the case of jointly Gaussian distributed sources, we establish an expression for the rate-distortion function under the constraint of quadratic distortion. Using the proposed rate-distortion function, any corner point on the rate-distortion region can be conveniently calculated. We take Wyner's common information as an example and provide a general and simple method to solve this problem. Through the analysis of the rate-distortion function, the rate on each layer and covariance matrix of auxiliary random variables and the sources are also presented in this paper.

Treatment of liver metastasis experiences slow progress owing to the severe side effects. In this study, we demonstrate a strategy capable of eliminating metastatic cancer cells in a selective manner. Nucleus-targeting W18O49 nanoparticles (WONPs) are conjugated to mitochondria-selective mesoporous silica nanoparticles (MSNs) containing photosensitizer (Ce6) through a Cathepsin B-cleavable peptide. In hepatocytes, upon the laser irradiation, the generated singlet oxygen species are consumed by WONPs, in turn leading to the loss of their photothermally heating capacity, thereby sparing hepatocyte from thermal damage induced by the laser illumination. By contrast, in cancer cells, the cleaved peptide linker allows WONPs and MSNs to respectively target nucleus and mitochondria, where the therapeutic powers could be unleashed, both photodynamically and photothermally. This ensures the energy production of cancer cells can be abolished. We further assess the underlying molecular mechanism at both gene and protein levels to better understand the therapeutic outcome.

Treatment of liver metastasis experiences slow progress owing to the severe side effects. In this study, we demonstrate a strategy capable of eliminating metastatic cancer cells in a selective manner. Nucleus-targeting W18O49 nanoparticles (WONPs) are conjugated to mitochondria-selective mesoporous silica nanoparticles (MSNs) containing photosensitizer (Ce6) through a Cathepsin B-cleavable peptide. In hepatocytes, upon the laser irradiation, the generated singlet oxygen species are consumed by WONPs, in turn leading to the loss of their photothermally heating capacity, thereby sparing hepatocyte from thermal damage induced by the laser illumination. By contrast, in cancer cells, the cleaved peptide linker allows WONPs and MSNs to respectively target nucleus and mitochondria, where the therapeutic powers could be unleashed, both photodynamically and photothermally. This ensures the energy production of cancer cells can be abolished. We further assess the underlying molecular mechanism at both gene and protein levels to better understand the therapeutic outcome.

Using precise times of every spike, spiking supervised learning has more effects on complex spatial-temporal pattern than supervised learning only through neuronal firing rates. The purpose of spiking supervised learning after spatial-temporal encoding is to emit desired spike trains with precise times. Existing algorithms of spiking supervised learning have excellent performances, but mechanisms of them still have some problems, such as the limitation of neuronal types and complex computation. Based on an online regulative mechanism of biological synapses, this paper proposes an online supervised learning algorithm of multiple spike trains for spiking neural networks. The proposed algorithm with a spatial-temporal transformation can make a simple direct regulation of synaptic weights as soon as firing time of an output spike is obtained. Besides, it is also not restricted by types of spiking neuron models. Relationship among desired output, actual output and input spike trains is firstly analyzed and synthesized to simply select a unit of pair-spike for a direct regulation. And then a computational method is constructed based on simple triple spikes using this direct regulation. Compared with other learning algorithms, results of experiments show that proposed algorithm has higher learning accuracy and efficiency., Comment: 11 pages, 5 figures and 5 tables, partly

Guojun Chen,1,* Longyuan Jiang,1,* Liwen Wang,1 Wei Zhang,2 Carlos Castillo,3,4 Xiangshao Fang1 1Department of Emergency Medicine, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China; 2Department of Respiratory Medicine, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China; 3Zhang Neuroscience Research Laboratories, School of Medicine, Loma Linda University, Loma Linda, CA, USA; 4Bioengineering Department, Gordon and Jill Bourns College of Engineering, California Baptist University, Riverside, CA, USA *These authors contributed equally to this work Background and aim: It is desirable to facilitate the use of an affordable, reliable, and portable spirometer, for earlier diagnosis of COPD in China, particularly in rural areas. The aim of this study was to assess the agreement of a handheld “disposable pneumotachograph” (D-PNEU) spirometer with the gold standard spirometer and to evaluate its diagnostic accuracy of spirometric classification of airflow obstruction. Subjects and methods: A total of 241 adult Chinese subjects ranging from healthy to those with mixed levels of pulmonary disease performed spirometry in a conventional body plethysmograph, and using a D-PNEU device in randomized order. The three best spirometric tests were recorded for comparative analysis. A Bland–Altman graph was created to assess the agreement between devices. Using FEV1/FVC <70% as the “gold standard” for obstruction, the accuracy of classifying the severity of airway obstruction for all subjects was assessed. For the specific individuals (n=159) able to exhale for at least 6 seconds, the accuracy of classifying airway obstruction was further assessed. For this purpose, a receiver operating characteristic curve was used to determine an optimal cutoff point of FEV1/FEV6 ratio obtained by the D-PNEU device, which matched the global definition of FEV1/FVC <70% by the traditional spirometer

In this paper, we present RKD-SLAM, a robust keyframe-based dense SLAM approach for an RGB-D camera that can robustly handle fast motion and dense loop closure, and run without time limitation in a moderate size scene. It not only can be used to scan high-quality 3D models, but also can satisfy the demand of VR and AR applications. First, we combine color and depth information to construct a very fast keyframe-based tracking method on a CPU, which can work robustly in challenging cases (e.g.~fast camera motion and complex loops). For reducing accumulation error, we also introduce a very efficient incremental bundle adjustment (BA) algorithm, which can greatly save unnecessary computation and perform local and global BA in a unified optimization framework. An efficient keyframe-based depth representation and fusion method is proposed to generate and timely update the dense 3D surface with online correction according to the refined camera poses of keyframes through BA. The experimental results and comparisons on a variety of challenging datasets and TUM RGB-D benchmark demonstrate the effectiveness of the proposed system., Comment: 12 pages, 9 figures

Background Enzymatic hydrolysis of lignocellulosic biomass (mainly plant cell walls) is a critical process for biofuel production. This process is greatly hindered by the natural complexity of plant cell walls and limited accessibility of surface cellulose by enzymes. Little is known about the plant cell wall structural and molecular level component changes after pretreatments, especially on the outer surface. Therefore, a more profound understanding of surface cellulose distributions before and after pretreatments at single-molecule level is in great need. In this study, we determined the structural changes, specifically on crystalline cellulose, of natural, dilute sulfuric acid pretreated and delignified cell wall surfaces of poplar, switchgrass, and corn stover using single molecular atomic force microscopy (AFM) recognition imaging. Results The AFM tip was first functionalized by a family 3 carbohydrate-binding module (CBM3a) (Clostridium thermocellum Scaffoldin) which specifically recognizes crystalline cellulose by selectively binding to it. The surface structural changes were studied at single molecule level based on the recognition area percentage (RAP) of exposed crystalline cellulose over the imaged cell wall surface. Our results show that the cell wall surface crystalline cellulose coverage increased from 17-20% to 18-40% after dilute acid pretreatment at 135°C under different acid concentrations and reached to 40-70% after delignification. Pretreated with 0.5% sulfuric acid, the crystalline cellulose surface distributions of 23% on poplar, 28% on switchgrass and, 38% on corn stover were determined as an optimized result. Corn stover cell walls also show less recalcitrance due to more effective pretreatments and delignification compared to poplar and switchgrass. Conclusions The dilute acid pretreatment can effectively increase the cellulose accessibility on plant cell wall surfaces. The optimal acid concentration was determined to be 0.5% acid a

Background Enzymatic hydrolysis of lignocellulosic biomass (mainly plant cell walls) is a critical process for biofuel production. This process is greatly hindered by the natural complexity of plant cell walls and limited accessibility of surface cellulose by enzymes. Little is known about the plant cell wall structural and molecular level component changes after pretreatments, especially on the outer surface. Therefore, a more profound understanding of surface cellulose distributions before and after pretreatments at single-molecule level is in great need. In this study, we determined the structural changes, specifically on crystalline cellulose, of natural, dilute sulfuric acid pretreated and delignified cell wall surfaces of poplar, switchgrass, and corn stover using single molecular atomic force microscopy (AFM) recognition imaging. Results The AFM tip was first functionalized by a family 3 carbohydrate-binding module (CBM3a) (Clostridium thermocellum Scaffoldin) which specifically recognizes crystalline cellulose by selectively binding to it. The surface structural changes were studied at single molecule level based on the recognition area percentage (RAP) of exposed crystalline cellulose over the imaged cell wall surface. Our results show that the cell wall surface crystalline cellulose coverage increased from 17-20% to 18-40% after dilute acid pretreatment at 135°C under different acid concentrations and reached to 40-70% after delignification. Pretreated with 0.5% sulfuric acid, the crystalline cellulose surface distributions of 23% on poplar, 28% on switchgrass and, 38% on corn stover were determined as an optimized result. Corn stover cell walls also show less recalcitrance due to more effective pretreatments and delignification compared to poplar and switchgrass. Conclusions The dilute acid pretreatment can effectively increase the cellulose accessibility on plant cell wall surfaces. The optimal acid concentration was determined to be 0.5% acid a