Search

Defect segmentation is crucial for quality control in advanced manufacturing, yet data scarcity poses challenges for state-of-the-art supervised deep learning. Synthetic defect data generation is a popular approach for mitigating data challenges. However, many current methods simply generate defects following a fixed set of rules, which may not directly relate to downstream task performance. This can lead to suboptimal performance and may even hinder the downstream task. To solve this problem, we leverage a novel bi-level optimization-based synthetic defect data generation framework. We use an online synthetic defect generation module grounded in the commonly-used Cut\&Paste framework, and adopt an efficient gradient-based optimization algorithm to solve the bi-level optimization problem. We achieve simultaneous training of the defect segmentation network, and learn various parameters of the data synthesis module by maximizing the validation performance of the trained defect segmentation network. Our experimental results on benchmark datasets under limited data settings show that the proposed bi-level optimization method can be used for learning the most effective locations for pasting synthetic defects thereby improving the segmentation performance by up to 18.3\% when compared to pasting defects at random locations. We also demonstrate up to 2.6\% performance gain by learning the importance weights for different augmentation-specific defect data sources when compared to giving equal importance to all the data sources.

The rapid development of large language models (LLMs) has significantly transformed the field of artificial intelligence, demonstrating remarkable capabilities in natural language processing and moving towards multi-modal functionality. These models are increasingly integrated into diverse applications, impacting both research and industry. However, their development and deployment present substantial challenges, including the need for extensive computational resources, high energy consumption, and complex software optimizations. Unlike traditional deep learning systems, LLMs require unique optimization strategies for training and inference, focusing on system-level efficiency. This paper surveys hardware and software co-design approaches specifically tailored to address the unique characteristics and constraints of large language models. This survey analyzes the challenges and impacts of LLMs on hardware and algorithm research, exploring algorithm optimization, hardware design, and system-level innovations. It aims to provide a comprehensive understanding of the trade-offs and considerations in LLM-centric computing systems, guiding future advancements in AI. Finally, we summarize the existing efforts in this space and outline future directions toward realizing production-grade co-design methodologies for the next generation of large language models and AI systems., Comment: Accepted by IEEE Circuits and Systems Magazine

With the growing popularity of RAG, the capabilities of embedding models are gaining increasing attention. Embedding models are primarily trained through contrastive loss learning, with negative examples being a key component. Previous work has proposed various hard negative mining strategies, but these strategies are typically employed as preprocessing steps. In this paper, we propose the conan-embedding model, which maximizes the utilization of more and higher-quality negative examples. Specifically, since the model's ability to handle preprocessed negative examples evolves during training, we propose dynamic hard negative mining method to expose the model to more challenging negative examples throughout the training process. Secondly, contrastive learning requires as many negative examples as possible but is limited by GPU memory constraints. Therefore, we use a Cross-GPU balancing Loss to provide more negative examples for embedding training and balance the batch size across multiple tasks. Moreover, we also discovered that the prompt-response pairs from LLMs can be used for embedding training. Our approach effectively enhances the capabilities of embedding models, currently ranking first on the Chinese leaderboard of Massive text embedding benchmark

Multimodal large language models (MLLMs) enhance the capabilities of standard large language models by integrating and processing data from multiple modalities, including text, vision, audio, video, and 3D environments. Data plays a pivotal role in the development and refinement of these models. In this survey, we comprehensively review the literature on MLLMs from a data-centric perspective. Specifically, we explore methods for preparing multimodal data during the pretraining and adaptation phases of MLLMs. Additionally, we analyze the evaluation methods for the datasets and review the benchmarks for evaluating MLLMs. Our survey also outlines potential future research directions. This work aims to provide researchers with a detailed understanding of the data-driven aspects of MLLMs, fostering further exploration and innovation in this field.

Modern diffusion-based image generative models have made significant progress and become promising to enrich training data for the object detection task. However, the generation quality and the controllability for complex scenes containing multi-class objects and dense objects with occlusions remain limited. This paper presents ODGEN, a novel method to generate high-quality images conditioned on bounding boxes, thereby facilitating data synthesis for object detection. Given a domain-specific object detection dataset, we first fine-tune a pre-trained diffusion model on both cropped foreground objects and entire images to fit target distributions. Then we propose to control the diffusion model using synthesized visual prompts with spatial constraints and object-wise textual descriptions. ODGEN exhibits robustness in handling complex scenes and specific domains. Further, we design a dataset synthesis pipeline to evaluate ODGEN on 7 domain-specific benchmarks to demonstrate its effectiveness. Adding training data generated by ODGEN improves up to 25.3% mAP@.50:.95 with object detectors like YOLOv5 and YOLOv7, outperforming prior controllable generative methods. In addition, we design an evaluation protocol based on COCO-2014 to validate ODGEN in general domains and observe an advantage up to 5.6% in mAP@.50:.95 against existing methods., Comment: Accepted by NeurIPS2024

Exploring dense connectivity of convolutional operators establishes critical "synapses" to communicate feature vectors from different levels and enriches the set of transformations on Computer Vision applications. Yet, even with heavy-machinery approaches such as Neural Architecture Search (NAS), discovering effective connectivity patterns requires tremendous efforts due to either constrained connectivity design space or a sub-optimal exploration process induced by an unconstrained search space. In this paper, we propose CSCO, a novel paradigm that fabricates effective connectivity of convolutional operators with minimal utilization of existing design motifs and further utilizes the discovered wiring to construct high-performing ConvNets. CSCO guides the exploration via a neural predictor as a surrogate of the ground-truth performance. We introduce Graph Isomorphism as data augmentation to improve sample efficiency and propose a Metropolis-Hastings Evolutionary Search (MH-ES) to evade locally optimal architectures and advance search quality. Results on ImageNet show ~0.6% performance improvement over hand-crafted and NAS-crafted dense connectivity. Our code is publicly available., Comment: To appear on Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops (2024)

One-dimensional (1D) interacting electrons are often described as a Luttinger liquid1-4 having properties that are intrinsically different from Fermi liquids in higher dimensions5,6. 1D electrons in materials systems exhibit exotic quantum phenomena that can be tuned by both intra- and inter-1D-chain electronic interactions, but their experimental characterization can be challenging. Here we demonstrate that layer-stacking domain walls (DWs) in van der Waals heterostructures form a broadly tunable Luttinger liquid system including both isolated and coupled arrays. We have imaged the evolution of DW Luttinger liquids under different interaction regimes tuned by electron density using a novel scanning tunneling microscopy (STM) technique. Single DWs at low carrier density are highly susceptible to Wigner crystallization consistent with a spin-incoherent Luttinger liquid, while at intermediate densities dimerized Wigner crystals form due to an enhanced magneto-elastic coupling. Periodic arrays of DWs exhibit an interplay between intra- and inter-chain interactions that gives rise to new quantum phases. At low electron densities inter-chain interactions are dominant and induce a 2D electron crystal composed of phased-locked 1D Wigner crystal in a staggered configuration. Increased electron density causes intra-chain fluctuation potentials to dominate, leading to an electronic smectic liquid crystal phase where electrons are ordered with algebraical correlation decay along the chain direction but disordered between chains. Our work shows that layer-stacking DWs in 2D heterostructures offers new opportunities to explore Luttinger liquid physics.

In medical and industrial domains, providing guidance for assembly processes can be critical to ensure efficiency and safety. Errors in assembly can lead to significant consequences such as extended surgery times and prolonged manufacturing or maintenance times in industry. Assembly scenarios can benefit from in-situ augmented reality visualization, i.e., augmentations in close proximity to the target object, to provide guidance, reduce assembly times, and minimize errors. In order to enable in-situ visualization, 6D pose estimation can be leveraged to identify the correct location for an augmentation. Existing 6D pose estimation techniques primarily focus on individual objects and static captures. However, assembly scenarios have various dynamics, including occlusion during assembly and dynamics in the appearance of assembly objects. Existing work focus either on object detection combined with state detection, or focus purely on the pose estimation. To address the challenges of 6D pose estimation in combination with assembly state detection, our approach ASDF builds upon the strengths of YOLOv8, a real-time capable object detection framework. We extend this framework, refine the object pose, and fuse pose knowledge with network-detected pose information. Utilizing our late fusion in our Pose2State module results in refined 6D pose estimation and assembly state detection. By combining both pose and state information, our Pose2State module predicts the final assembly state with precision. The evaluation of our ASDF dataset shows that our Pose2State module leads to an improved assembly state detection and that the improvement of the assembly state further leads to a more robust 6D pose estimation. Moreover, on the GBOT dataset, we outperform the pure deep learning-based network and even outperform the hybrid and pure tracking-based approaches.

Acoustic streaming shows great potential in applications such as bubble dynamics, cell aggregation, and nano-sized particle isolation in the biomedical and drug industries. As the acoustic shock distance decreases with the increase of incident frequency, the nonlinear propagation effect will play a role in acoustic streaming, e.g., Eckart (bulk) streaming at a few gigahertz (GHz). However, the theory of source terms of bulk streaming is still missing at this stage when high-order acoustic harmonics play a role. In this paper, we derive the source term including the contribution of higher-order harmonics. The streaming-induced hydrodynamic flow is assumed to be incompressible and no shock wave occurs during the nonlinear acoustic propagation as restricted by the traditional Goldberg number \Gamma < 1 or \Gamma \approx 1 which indicates the importance of nonlinearity relative to dissipation. The derived force terms allow evaluating bulk streaming with high-order harmonics at GHz and provide an exact expression compared to the existing empirical formulas. Numerical results show that the contribution of higher-order harmonics increases the streaming flow velocity by more than 20%. We show that the expression introduced by Nyborg should be avoided in numerical computations as it includes part of the acoustic radiation force that does not lead to acoustic streaming., Comment: 11 pages, 7 figures

Guidance for assemblable parts is a promising field for augmented reality. Augmented reality assembly guidance requires 6D object poses of target objects in real time. Especially in time-critical medical or industrial settings, continuous and markerless tracking of individual parts is essential to visualize instructions superimposed on or next to the target object parts. In this regard, occlusions by the user's hand or other objects and the complexity of different assembly states complicate robust and real-time markerless multi-object tracking. To address this problem, we present Graph-based Object Tracking (GBOT), a novel graph-based single-view RGB-D tracking approach. The real-time markerless multi-object tracking is initialized via 6D pose estimation and updates the graph-based assembly poses. The tracking through various assembly states is achieved by our novel multi-state assembly graph. We update the multi-state assembly graph by utilizing the relative poses of the individual assembly parts. Linking the individual objects in this graph enables more robust object tracking during the assembly process. For evaluation, we introduce a synthetic dataset of publicly available and 3D printable assembly assets as a benchmark for future work. Quantitative experiments in synthetic data and further qualitative study in real test data show that GBOT can outperform existing work towards enabling context-aware augmented reality assembly guidance. Dataset and code will be made publically available., Comment: 9 pages

The behavior of two-dimensional electron gas (2DEG) in extreme coupling limits are reasonably well-understood, but our understanding of intermediate region remains limited. Strongly interacting electrons crystalize into a solid phase known as the Wigner crystal at very low densities, and these evolve to a Fermi liquid at high densities. At intermediate densities, however, where the Wigner crystal melts into a strongly correlated electron fluid that is poorly understood partly due to a lack of microscopic probes for delicate quantum phases. Here we report the first imaging of a disordered Wigner solid and its quantum densification and quantum melting behavior in a bilayer MoSe2 using a non-invasive scanning tunneling microscopy (STM) technique. We observe a Wigner solid with nanocrystalline domains pinned by local disorder at low hole densities. With slightly increasing electrostatic gate voltages, the holes are added quantum mechanically during the densification of the disordered Wigner solid. As the hole density is increased above a threshold (p ~ 5.7 * 10e12 (cm-2)), the Wigner solid is observed to melt locally and create a mixed phase where solid and liquid regions coexist. With increasing density, the liquid regions gradually expand and form an apparent percolation network. Local solid domains appear to be pinned and stabilized by local disorder over a range of densities. Our observations are consistent with a microemulsion picture of Wigner solid quantum melting where solid and liquid domains emerge spontaneously and solid domains are pinned by local disorder.

Background: Type 2 diabetes (T2D) is a prevalent chronic disease with a significant risk of serious health complications and negative impacts on the quality of life. Given the impact of individual characteristics and lifestyle on the treatment plan and patient outcomes, it is crucial to develop precise and personalized management strategies. Artificial intelligence (AI) provides great promise in combining patterns from various data sources with nurses' expertise to achieve optimal care. Methods: This is a 6-month ancillary study among T2D patients (n = 20, age = 57 +- 10). Participants were randomly assigned to an intervention (AI, n=10) group to receive daily AI-generated individualized feedback or a control group without receiving the daily feedback (non-AI, n=10) in the last three months. The study developed an online nurse-in-the-loop predictive control (ONLC) model that utilizes a predictive digital twin (PDT). The PDT was developed using a transfer-learning-based Artificial Neural Network. The PDT was trained on participants self-monitoring data (weight, food logs, physical activity, glucose) from the first three months, and the online control algorithm applied particle swarm optimization to identify impactful behavioral changes for maintaining the patient's glucose and weight levels for the next three months. The ONLC provided the intervention group with individualized feedback and recommendations via text messages. The PDT was re-trained weekly to improve its performance. Findings: The trained ONLC model achieved >=80% prediction accuracy across all patients while the model was tuned online. Participants in the intervention group exhibited a trend of improved daily steps and stable or improved total caloric and total carb intake as recommended., Comment: Submitted for review

Power efficiency is a critical design objective in modern microprocessor design. To evaluate the impact of architectural-level design decisions, an accurate yet efficient architecture-level power model is desired. However, widely adopted data-independent analytical power models like McPAT and Wattch have been criticized for their unreliable accuracy. While some machine learning (ML) methods have been proposed for architecture-level power modeling, they rely on sufficient known designs for training and perform poorly when the number of available designs is limited, which is typically the case in realistic scenarios. In this work, we derive a general formulation that unifies existing architecture-level power models. Based on the formulation, we propose PANDA, an innovative architecture-level solution that combines the advantages of analytical and ML power models. It achieves unprecedented high accuracy on unknown new designs even when there are very limited designs for training, which is a common challenge in practice. Besides being an excellent power model, it can predict area, performance, and energy accurately. PANDA further supports power prediction for unknown new technology nodes. In our experiments, besides validating the superior performance and the wide range of functionalities of PANDA, we also propose an application scenario, where PANDA proves to identify high-performance design configurations given a power constraint.

Approximate nearest neighbor search (ANNS) is a key retrieval technique for vector database and many data center applications, such as person re-identification and recommendation systems. It is also fundamental to retrieval augmented generation (RAG) for large language models (LLM) now. Among all the ANNS algorithms, graph-traversal-based ANNS achieves the highest recall rate. However, as the size of dataset increases, the graph may require hundreds of gigabytes of memory, exceeding the main memory capacity of a single workstation node. Although we can do partitioning and use solid-state drive (SSD) as the backing storage, the limited SSD I/O bandwidth severely degrades the performance of the system. To address this challenge, we present NDSEARCH, a hardware-software co-designed near-data processing (NDP) solution for ANNS processing. NDSEARCH consists of a novel in-storage computing architecture, namely, SEARSSD, that supports the ANNS kernels and leverages logic unit (LUN)-level parallelism inside the NAND flash chips. NDSEARCH also includes a processing model that is customized for NDP and cooperates with SEARSSD. The processing model enables us to apply a two-level scheduling to improve the data locality and exploit the internal bandwidth in NDSEARCH, and a speculative searching mechanism to further accelerate the ANNS workload. Our results show that NDSEARCH improves the throughput by up to 31.7x, 14.6x, 7.4x 2.9x over CPU, GPU, a state-of-the-art SmartSSD-only design, and DeepStore, respectively. NDSEARCH also achieves two orders-of-magnitude higher energy efficiency than CPU and GPU.

Theoretical bounds are commonly used to assess the limitations of photonic design. Here we introduce a more active way to use theoretical bounds, integrating them into part of the design process and identifying optimal system parameters that maximize the efficiency limit itself. As an example, we consider wide-field-of-view high-numerical-aperture metalenses, which can be used for high-resolution imaging in microscopy and endoscopy, but no existing design has achieved a high efficiency. By choosing aperture sizes to maximize an efficiency bound, setting the thickness according to a thickness bound, and then performing inverse design, we come up with high-numerical-aperture (NA = 0.9) metalens designs with record-high 98% transmission efficiency and 92% Strehl ratio across all incident angles within a 60-deg field of view, reaching the maximized bound. This maximizing-efficiency-limit approach applies to any multi-channel system and can help a wide range of optical devices reach their highest possible performance.

Mixture-of-Experts (MoE) has emerged as a favorable architecture in the era of large models due to its inherent advantage, i.e., enlarging model capacity without incurring notable computational overhead. Yet, the realization of such benefits often results in ineffective GPU memory utilization, as large portions of the model parameters remain dormant during inference. Moreover, the memory demands of large models consistently outpace the memory capacity of contemporary GPUs. Addressing this, we introduce SiDA-MoE ($\textbf{S}$parsity-$\textbf{i}$nspired $\textbf{D}$ata-$\textbf{A}$ware), an efficient inference approach tailored for large MoE models. SiDA-MoE judiciously exploits both the system's main memory, which is now abundant and readily scalable, and GPU memory by capitalizing on the inherent sparsity on expert activation in MoE models. By adopting a data-aware perspective, SiDA-MoE achieves enhanced model efficiency with a neglectable performance drop. Specifically, SiDA-MoE attains a remarkable speedup in MoE inference with up to $3.93\times$ throughput increasing, up to $72\%$ latency reduction, and up to $80\%$ GPU memory saving with down to $1\%$ performance drop. This work paves the way for scalable and efficient deployment of large MoE models, even with constrained resources. Code is available at: https://github.com/timlee0212/SiDA-MoE., Comment: Published on MLSys24. https://openreview.net/forum?id=q26ydTFF5j}

Resistive random access memory (ReRAM)-based processing-in-memory (PIM) architectures have demonstrated great potential to accelerate Deep Neural Network (DNN) training/inference. However, the computational accuracy of analog PIM is compromised due to the non-idealities, such as the conductance variation of ReRAM cells. The impact of these non-idealities worsens as the number of concurrently activated wordlines and bitlines increases. To guarantee computational accuracy, only a limited number of wordlines and bitlines of the crossbar array can be turned on concurrently, significantly reducing the achievable parallelism of the architecture. While the constraints on parallelism limit the efficiency of the accelerators, they also provide a new opportunity for fine-grained mixed-precision quantization. To enable efficient DNN inference on practical ReRAM-based accelerators, we propose an algorithm-architecture co-design framework called \underline{B}lock-\underline{W}ise mixed-precision \underline{Q}uantization (BWQ). At the algorithm level, BWQ-A introduces a mixed-precision quantization scheme at the block level, which achieves a high weight and activation compression ratio with negligible accuracy degradation. We also present the hardware architecture design BWQ-H, which leverages the low-bit-width models achieved by BWQ-A to perform high-efficiency DNN inference on ReRAM devices. BWQ-H also adopts a novel precision-aware weight mapping method to increase the ReRAM crossbar's throughput. Our evaluation demonstrates the effectiveness of BWQ, which achieves a 6.08x speedup and a 17.47x energy saving on average compared to existing ReRAM-based architectures., Comment: 12 pages, 13 figures

Amorphous materials exhibit various characteristics that are not featured by crystals and can sometimes be tuned by their degree of disorder (DOD). Here, we report results on the mechanical properties of monolayer amorphous carbon (MAC) and monolayer amorphous boron nitride (maBN) with different DOD. The pertinent structures are obtained by kinetic-Monte-Carlo (kMC) simulations using machine-learning potentials (MLP) with density-functional-theory (DFT)-level accuracy. An intuitive order parameter, namely the areal fraction Fx occupied by crystallites within the continuous random network, is proposed to describe the DOD. We find that Fx captures the essence of the DOD: Samples with the same Fx but different sizes and distributions of crystallites have virtually identical radial distributions functions as well as bond-length and bond-angle distributions. Furthermore, by simulating the fracture process with molecular dynamics, we found that the mechanical responses of MAC and maBN before fracture are solely determined by Fx and are insensitive to the sizes and specific arrangements of the crystallites. The behavior of cracks in the two materials is analyzed and found to mainly propagate in meandering paths in the CRN region and to be influenced by crystallites in distinct ways that toughen the material. The present results reveal the relation between structure and mechanical properties in amorphous monolayers and may provide a universal toughening strategy for 2D materials.

Computer-automated design and discovery have led to high-performance nanophotonic devices with diverse functionalities. However, massively multi-channel systems such as metasurfaces controlling many incident angles and photonic-circuit components coupling many waveguide modes still present a challenge. Conventional methods require $M_{\rm in}$ forward simulations and $M_{\rm in}$ adjoint simulations -- $2M_{\rm in}$ simulations in total -- to compute the objective function and its gradient for a design involving the response to $M_{\rm in}$ input channels. By generalizing the adjoint method and the recently proposed augmented partial factorization method, here we show how to obtain both the objective function and its gradient for a massively multi-channel system in a single simulation, achieving over-two-orders-of-magnitude speedup and reduced memory usage. We use this method to inverse design a metasurface beam splitter that separates the incident light to the target diffraction orders for all incident angles of interest, a key component of the dot projector for 3D sensing. This formalism enables efficient inverse design for a wide range of multi-channel optical systems.

The rapidly advancing capabilities in nanophotonic design are enabling complex functionalities limited mainly by physical bounds. The efficiency of transmission is a major consideration, but its ultimate limit remains unknown for most systems. Here, we introduce a matrix formalism that puts a fundamental bound on the channel-averaged transmission efficiency of any passive multi-channel optical system based only on energy conservation and the desired functionality, independent of the interior structure and material composition. Applying this formalism to diffraction-limited nonlocal metalenses with a wide field of view, we show that the transmission efficiency must decrease with the numerical aperture for the commonly adopted designs with equal entrance and output aperture diameters. We also show that reducing the size of the entrance aperture can raise the efficiency bound. This work reveals a fundamental limit on the transmission efficiency as well as providing guidance for the design of high-efficiency multi-channel optical systems.

Visual anomaly detection, an important problem in computer vision, is usually formulated as a one-class classification and segmentation task. The student-teacher (S-T) framework has proved to be effective in solving this challenge. However, previous works based on S-T only empirically applied constraints on normal data and fused multi-level information. In this study, we propose an improved model called DeSTSeg, which integrates a pre-trained teacher network, a denoising student encoder-decoder, and a segmentation network into one framework. First, to strengthen the constraints on anomalous data, we introduce a denoising procedure that allows the student network to learn more robust representations. From synthetically corrupted normal images, we train the student network to match the teacher network feature of the same images without corruption. Second, to fuse the multi-level S-T features adaptively, we train a segmentation network with rich supervision from synthetic anomaly masks, achieving a substantial performance improvement. Experiments on the industrial inspection benchmark dataset demonstrate that our method achieves state-of-the-art performance, 98.6% on image-level AUC, 75.8% on pixel-level average precision, and 76.4% on instance-level average precision., Comment: Accepted by CVPR 2023

Metalenses -- flat lenses made with optical metasurfaces -- promise to enable thinner, cheaper, and better imaging systems. Achieving a sufficient angular field of view (FOV) is crucial toward that goal and requires a tailored incident-angle-dependent response. Here, we show that there is an intrinsic trade-off between achieving a desired broad-angle response and reducing the thickness of the device. It originates from the Fourier transform duality between space and angle. One can write down the transmission matrix describing the desired angle-dependent response, convert it to the spatial basis where its degree of nonlocality can be quantified through a lateral spreading, and determine the minimal device thickness based on such a required lateral spreading. This approach is general. When applied to wide-FOV lenses, it predicts the minimal thickness as a function of the FOV, lens diameter, and numerical aperture. The bound is tight, as some inverse-designed multi-layer metasurfaces can approach the minimal thickness we found. This work offers guidance for the design of nonlocal metasurfaces, proposes a new framework for establishing bounds, and reveals the relation between angular diversity and spatial footprint in multi-channel systems.

Image beauty assessment is an important subject of computer vision. Therefore, building a model to mimic the image beauty assessment becomes an important task. To better imitate the behaviours of the human visual system (HVS), a complete survey about images of different categories should be implemented. This work focuses on image beauty assessment. In this study, the pairwise evaluation method was used, which is based on the Bradley-Terry model. We believe that this method is more accurate than other image rating methods within an image group. Additionally, Convolution neural network (CNN), which is fit for image quality assessment, is used in this work. The first part of this study is a survey about the image beauty comparison of different images. The Bradley-Terry model is used for the calculated scores, which are the target of CNN model. The second part of this work focuses on the results of the image beauty prediction, including landscape images, architecture images and portrait images. The models are pretrained by the AVA dataset to improve the performance later. Then, the CNN model is trained with the surveyed images and corresponding scores. Furthermore, this work compares the results of four CNN base networks, i.e., Alex net, VGG net, Squeeze net and LSiM net, as discussed in literature. In the end, the model is evaluated by the accuracy in pairs, correlation coefficient and relative error calculated by survey results. Satisfactory results are achieved by our proposed methods with about 70 percent accuracy in pairs. Our work sheds more light on the novel image beauty assessment method. While more studies should be conducted, this method is a promising step., Comment: 16 pages 18 fiugures