Pruning and quantization for deep neural network acceleration: A survey

Agriculture plays a significant role in the growth and development of any nation’s economy. But, the emergence of several crop-related diseases affects the productivity in the agriculture sector. To cope up this issue and to make aware the farmers to prevent the expansion of diseases in crops and to implement effective management, crop disease diagnosis plays its significant role. Researchers had already used many techniques for this purpose, but some vision-related techniques are yet to be explored. Commonly used techniques are support vector machine, k-means clustering, radial basis functions, genetic algorithm, image processing techniques like filtering and segmentation, deep structured learning techniques like convolutional neural network. We have designed a hybrid approach for detection of crop leaf diseases using the combination of convolutional neural networks and autoencoders. This research paper provides a novel technique to detect crop diseases with the help of convolutional encoder networks using crop leaf images. We have obtained our result over a 900-image dataset, out of which 600 constitute the training set and 300 test set. We have considered 3 crops and 5 kinds of crop diseases. The proposed network was trained in such a way that it can distinguish the crop disease using the leaf images. Different convolution filters like 2 × 2 and 3 × 3 are used in proposed work. It was observed that the proposed architecture achieved variation in accuracy for the different number of epochs and for different convolution filter size. We reached 97.50% accuracy for 2 × 2 convolution filter size in 100 epochs, while 100% accuracy for 3 × 3 filter size which is better than other conventional methods.

Seasonal Crops Disease Prediction and Classification Using Deep Convolutional Encoder Network

Unsupervised domain adaptation aims to leverage the labeled source data to learn with the unlabeled target data. Previous trandusctive methods tackle it by iteratively seeking a low-dimensional projection to extract the invariant features and obtaining the pseudo target labels via building a classifier on source data. However, they merely concentrate on minimizing the cross-domain distribution divergence, while ignoring the intra-domain structure especially for the target domain. Even after projection, possible risk factors like imbalanced data distribution may still hinder the performance of target label inference. In this paper, we propose a simple yet effective domain-invariant projection ensemble approach to tackle these two issues together. Specifically, we seek the optimal projection via a novel relaxed domain-irrelevant clustering-promoting term that jointly bridges the cross-domain semantic gap and increases the intra-class compactness in both domains. To further enhance the target label inference, we first develop a ‘sampling-and-fusion’ framework, under which multiple projections are independently learned based on various randomized coupled domain subsets. Subsequently, aggregating models such as majority voting are utilized to leverage multiple projections and classify unlabeled target data. Extensive experimental results on six visual benchmarks including object, face, and digit images, demonstrate that the proposed methods gain remarkable margins over state-of-the-art unsupervised domain adaptation methods.

Aggregating Randomized Clustering-Promoting Invariant Projections for Domain Adaptation

The Internet of Things (IoTs) enables coupling of digital and physical objects using worthy communication technologies and introduces a future vision where computing systems, users and objects cooperate for convenience and economic benefits. Such a vision requires seamless security, data privacy, authentication and robustness against attacks. These attributes can be introduced by blockchain, a distributed ledger that maintains an immutable log of network transactions. In this paper, we present a comprehensive review on how to remodel blockchain to the specific IoT needs in order to develop Blockchain based IoT (BIoT) applications and aim to shape a coherent picture of the current state-of-the-art efforts in this direction. After describing the basic characteristics and requirements of IoT, evolution of blockchain is presented. In this regard, we start with the fundamental working principles of blockchain and how such systems achieve auditability, security and decentralization. Further, we describe the most relevant BIoT applications, its architecture design and security aspects. From there, we build our narrative on the centralized IoT challenges followed by recent advances towards solving them. Finally, some future directions are enumerated with the aim to guide future BIoT researchers on challenges that needs to be considered ahead of deploying the next generation of BIoT applications.

Unification of Blockchain and Internet of Things (BIoT): requirements, working model, challenges and future directions

Exploring uncertainty in pseudo-label guided unsupervised domain adaptation

Transfer learning (TL) aims at solving the problem of learning an effective classification model for a target category, which has few training samples, by leveraging knowledge from source categories with far more training data. We propose a new discriminative TL (DTL) method, combining a series of hypotheses made by both the model learned with target training samples and the additional models learned with source category samples. Specifically, we use the sparse reconstruction residual as a basic discriminant and enhance its discriminative power by comparing two residuals from a positive and a negative dictionary. On this basis, we make use of similarities and dissimilarities by choosing both positively correlated and negatively correlated source categories to form additional dictionaries. A new Wilcoxon–Mann–Whitney statistic-based cost function is proposed to choose the additional dictionaries with unbalanced training data. Also, two parallel boosting processes are applied to both the positive and negative data distributions to further improve classifier performance. On two different image classification databases, the proposed DTL consistently outperforms other state-of-the-art TL methods while at the same time maintaining very efficient runtime.

Discriminative Transfer Learning Using Similarities and Dissimilarities

Domain adaptation (DA) aims to generalize a learning model across training and testing data despite the mismatch of their data distributions. In light of a theoretical estimation of the upper error bound, we argue, in this article, that an effective DA method for classification should: 1) search a shared feature subspace where the source and target data are not only aligned in terms of distributions as most state-of-the-art DA methods do but also discriminative in that instances of different classes are well separated and 2) account for the geometric structure of the underlying data manifold when inferring data labels on the target domain. In comparison with a baseline DA method which only cares about data distribution alignment between source and target, we derive three different DA models for classification, namely, close yet discriminative DA (CDDA), geometry-aware DA (GA-DA), and discriminative and GA-DA (DGA-DA), to highlight the contribution of CDDA based on 1), GA-DA based on 2), and, finally, DGA-DA implementing jointly 1) and 2). Using both the synthetic and real data, we show the effectiveness of the proposed approach which consistently outperforms the state-of-the-art DA methods over 49 image classification DA tasks through eight popular benchmarks. We further carry out an in-depth analysis of the proposed DA method in quantifying the contribution of each term of our DA model and provide insights into the proposed DA methods in visualizing both real and synthetic data.

Discriminative and Geometry-Aware Unsupervised Domain Adaptation

We investigate the role that geometric, textual and visual features play in the task of predicting a preposition that links two visual entities depicted in an image. The task is an important part of the subsequent process of generating image descriptions. We explore the prediction of prepositions for a pair of entities, both in the case when the labels of such entities are known and unknown. In all situations we found clear evidence that all three features contribute to the prediction task.

/pdf/combining-geometric-textual-and-visual-features-for-2eeni6d68k.pdf

Combining Geometric, Textual and Visual Features for Predicting Prepositions in Image Descriptions

In this survey, we propose to explore and discuss the common rules behind knowledge transfer works for vision recognition tasks. To achieve this, we firstly discuss the different kinds of reusable knowledge existing in a vision recognition task, and then we categorize different knowledge transfer approaches depending on where the knowledge comes from and where the knowledge goes. Compared to previous surveys on knowledge transfer that are from the problem-oriented perspective or from the technique-oriented perspective, our viewpoint is closer to the nature of knowledge transfer and reveals common rules behind different transfer learning settings and applications. Besides different knowledge transfer categories, we also show some research works that study the transferability between different vision recognition tasks. We further give a discussion about the introduced research works and show some potential research directions in this field.

https://hal.archives-ouvertes.fr/hal-02101005/document

Knowledge Transfer in Vision Recognition: A Survey

Domain adaptation (DA) aims to generalize a learning model across training and testing data despite the mismatch of their data distributions. In light of a theoretical estimation of upper error bound, we argue in this paper that an effective DA method should 1) search a shared feature subspace where source and target data are not only aligned in terms of distributions as most state of the art DA methods do, but also discriminative in that instances of different classes are well separated; 2) account for the geometric structure of the underlying data manifold when inferring data labels on the target domain. In comparison with a baseline DA method which only cares about data distribution alignment between source and target, we derive three different DA models, namely CDDA, GA-DA, and DGA-DA, to highlight the contribution of Close yet Discriminative DA(CDDA) based on 1), Geometry Aware DA (GA-DA) based on 2), and finally Discriminative and Geometry Aware DA (DGA-DA) implementing jointly 1) and 2). Using both synthetic and real data, we show the effectiveness of the proposed approach which consistently outperforms state of the art DA methods over 36 image classification DA tasks through 6 popular benchmarks. We further carry out in-depth analysis of the proposed DA method in quantifying the contribution of each term of our DA model and provide insights into the proposed DA methods in visualizing both real and synthetic data.

Ying Lu

Papers

Discriminative Transfer Learning Using Similarities and Dissimilarities

Discriminative and Geometry-Aware Unsupervised Domain Adaptation

Combining Geometric, Textual and Visual Features for Predicting Prepositions in Image Descriptions

Knowledge Transfer in Vision Recognition: A Survey

Discriminative and Geometry Aware Unsupervised Domain Adaptation