Current home appliances are capable to execute a limited number of voice commands such as turning devices on or off, adjusting music volume or light conditions. Recent progress in machine reasoning gives an opportunity to develop new types of conversational user interfaces for home appliances. In this paper, we apply state-of-the-art visual reasoning model and demonstrate that it is feasible to ask a smart fridge about its contents and various properties of the food with close-to-natural conversation experience. Our visual reasoning model answers user questions about existence, count, category and freshness of each product by analyzing photos made by the image sensor inside the smart fridge. Users may chat with their fridge using off-the-shelf phone messenger while being away from home, for example, when shopping in the supermarket. We generate a visually realistic synthetic dataset to train machine learning reasoning model that achieves 95% answer accuracy on test data. We present the results of initial user tests and discuss how we modify distribution of generated questions for model training based on human-in-the-loop guidance. We open source code for the whole system including dataset generation, reasoning model and demonstration scripts.
Face is one of the most important things for communication with the world around us. It also forms our identity and expressions. Estimating the face structure is a fundamental task in computer vision with applications in different areas such as face recognition and medical surgeries. Recently, deep learning techniques achieved significant results for 3D face reconstruction from flat images. The main challenge of such techniques is a vital need for large 3D face datasets. Usually, this challenge is handled by synthetic face generation. However, synthetic datasets suffer from the existence of non-possible faces. Here, we propose a face manifold learning method for synthetic diverse face dataset generation. First, the face structure is divided into the shape and expression groups. Then, a fully convolutional autoencoder network is exploited to deal with the non-possible faces, and, simultaneously, preserving the dataset diversity. Simulation results show that the proposed method is capable of denoising highly corrupted faces. The diversity of the generated dataset is evaluated qualitatively and quantitatively and compared to the existing methods. Experiments show that our manifold learning method outperforms the state of the art methods significantly.
Training an object instance detector where only a few training object images are available is a challenging task. One solution is a cut-and-paste method that generates a training dataset by cutting object areas out of training images and pasting them onto other background images. A detector trained on a dataset generated with a cut-and-paste method suffers from the conventional domain shift problem, which stems from a discrepancy between the source domain (generated training dataset) and the target domain (real test dataset). Though state-of-the-art domain adaptation methods are able to reduce this gap, it is limited because they do not consider the difference of domain gaps of foreground and background. In this study, we present that the conventional domain gap can be divided into two sub-domain gaps for foreground and background. Then, we show that the original cut-and-paste approach suffers from a new domain gap problem, an unbalanced domain gaps, because it has two separate source domains for foreground and background, unlike the conventional domain shift problem. Then, we introduce an advanced cut-and-paste method to balance the unbalanced domain gaps by diversifying the foreground with GAN (generative adversarial network)-generated seed images and simplifying the background using image processing techniques. Experimental results show that our method is effective for balancing domain gaps and improving the accuracy of object instance detection in a cluttered indoor environment using only a few seed images. Furthermore, we show that balancing domain gaps can improve the detection accuracy of state-of-the-art domain adaptation methods.
The availability of large image data sets has been a crucial factor in the success of deep learning-based classification and detection methods. While data sets for everyday objects are widely available, data for specific industrial use-cases (e.g. identifying packaged products in a warehouse) remains scarce. In such cases, the data sets have to be created from scratch, placing a crucial bottleneck on the deployment of deep learning techniques in industrial applications. We present work carried out in collaboration with a leading UK online supermarket, with the aim of creating a computer vision system capable of detecting and identifying unique supermarket products in a warehouse setting. To this end, we demonstrate a framework for using synthetic data to create an end-to-end deep learning pipeline, beginning with real-world objects and culminating in a trained model. Our method is based on the generation of a synthetic dataset from 3D models obtained by applying photogrammetry techniques to real-world objects. Using 100k synthetic images generated from 60 real images per class, an InceptionV3 convolutional neural network (CNN) was trained, which achieved classification accuracy of 95.8% on a separately acquired test set of real supermarket product images. The image generation process supports automatic pixel annotation. This eliminates the prohibitively expensive manual annotation typically required for detection tasks. Based on this readily available data, a one-stage RetinaNet detector was trained on the synthetic, annotated images to produce a detector that can accurately localize and classify the specimen products in real-time.
Meta-learning extracts common knowledge from learning different tasks and uses it for unseen tasks. It can significantly improve tasks that suffer from insufficient training data, e.g., few shot learning. In most meta-learning methods, tasks are implicitly related by sharing parameters or optimizer. In this paper, we show that a meta-learner that explicitly relates tasks on a graph describing the relations of their output dimensions (e.g., classes) can significantly improve few shot learning. The graph's structure is usually free or cheap to obtain but has rarely been explored in previous works. We develop a novel meta-learner of this type for prototype-based classification, in which a prototype is generated for each class, such that the nearest neighbor search among the prototypes produces an accurate classification. The meta-learner, called "Gated Propagation Network (GPN)", learns to propagate messages between prototypes of different classes on the graph, so that learning the prototype of each class benefits from the data of other related classes. In GPN, an attention mechanism aggregates messages from neighboring classes of each class, with a gate choosing between the aggregated message and the message from the class itself. We train GPN on a sequence of tasks from many-shot to few shot generated by subgraph sampling. During training, it is able to reuse and update previously achieved prototypes from the memory in a life-long learning cycle. In experiments, under different training-test discrepancy and test task generation settings, GPN outperforms recent meta-learning methods on two benchmark datasets. The code of GPN and dataset generation is available at https://github.com/liulu112601/Gated-Propagation-Net.
Deep learning has achieved overwhelming success, spanning from discriminative models to generative models. In particular, deep generative models have facilitated a new level of performance in a myriad of areas, ranging from media manipulation to sanitized dataset generation. Despite the great success, the potential risks of privacy breach caused by generative models have not been analyzed systematically. In this paper, we focus on membership inference attack against deep generative models that reveals information about the training data used for victim models. Specifically, we present the first taxonomy of membership inference attacks, encompassing not only existing attacks but also our novel ones. In addition, we propose the first generic attack model that can be instantiated in a large range of settings and is applicable to various kinds of deep generative models. Moreover, we provide a theoretically grounded attack calibration technique, which consistently boosts the attack performance in all cases, across different attack settings, data modalities, and training configurations. We complement the systematic analysis of attack performance by a comprehensive experimental study, that investigates the effectiveness of various attacks w.r.t. model type and training configurations, over three diverse application scenarios (i.e., images, medical data, and location data).
The ever increasing volume of data in digital forensic investigation is one of the most discussed challenges in the field. Usually, most of the file artefacts on seized devices are not pertinent to the investigation. Manually retrieving suspicious files relevant to the investigation is akin to finding a needle in a haystack. In this paper, a methodology for the automatic prioritisation of suspicious file artefacts (i.e., file artefacts that are pertinent to the investigation) is proposed to reduce the manual analysis effort required. This methodology is designed to work in a human-in-the-loop fashion. In other words, it predicts/recommends that an artefact is likely to be suspicious rather than giving the final analysis result. A supervised machine learning approach is employed, which leverages the recorded results of previously processed cases. The process of features extraction, dataset generation, training and evaluation are presented in this paper. In addition, a toolkit for data extraction from disk images is outlined, which enables this method to be integrated with the conventional investigation process and work in an automated fashion.
In this technical report I present my method for automatic synthetic dataset generation for object detection and demonstrate it on the video game League of Legends. This report furthermore serves as a handbook on how to automatically generate datasets and as an introduction on the dataset generation part of the LeagueAI framework. The LeagueAI framework is a software framework that provides detailed information about the game League of Legends based on the same input a human player would have, namely vision. The framework allows researchers and enthusiasts to develop their own intelligent agents or to extract detailed information about the state of the game. A big problem of machine vision applications usually is the laborious work of gathering large amounts of hand labeled data. Thus, a crucial part of the vision pipeline of the LeagueAI framework, the dataset generation, is presented in this report. The method involves extracting image raw data from the game's 3D models and combining them with the game background to create game-like synthetic images and to generate the corresponding labels automatically. In an experiment I compared a model trained on synthetic data to a model trained on hand labeled data and a model trained on a combined dataset. The model trained on the synthetic data showed higher detection precision on more classes and more reliable tracking performance of the player character. The model trained on the combined dataset did not perform better because of the different formats of the older hand labeled dataset and the synthetic data.
Document structure analysis, such as zone segmentation and table recognition, is a complex problem in document processing and is an active area of research. The recent success of deep learning in solving various computer vision and machine learning problems has not been reflected in document structure analysis since conventional neural networks are not well suited to the input structure of the problem. In this paper, we propose an architecture based on graph networks as a better alternative to standard neural networks for table recognition. We argue that graph networks are a more natural choice for these problems, and explore two gradient-based graph neural networks. Our proposed architecture combines the benefits of convolutional neural networks for visual feature extraction and graph networks for dealing with the problem structure. We empirically demonstrate that our method outperforms the baseline by a significant margin. In addition, we identify the lack of large scale datasets as a major hindrance for deep learning research for structure analysis and present a new large scale synthetic dataset for the problem of table recognition. Finally, we open-source our implementation of dataset generation and the training framework of our graph networks to promote reproducible research in this direction.
In this paper, we propose a Seed-Augment-Train/Transfer (SAT) framework that contains a synthetic seed image dataset generation procedure for languages with different numeral systems using freely available open font file datasets. This seed dataset of images is then augmented to create a purely synthetic training dataset, which is in turn used to train a deep neural network and test on held-out real world handwritten digits dataset spanning five Indic scripts, Kannada, Tamil, Gujarati, Malayalam, and Devanagari. We showcase the efficacy of this approach both qualitatively, by training a Boundary-seeking GAN (BGAN) that generates realistic digit images in the five languages, and also quantitatively by testing a CNN trained on the synthetic data on the real-world datasets. This establishes not only an interesting nexus between the font-datasets-world and transfer learning but also provides a recipe for universal-digit classification in any script.
Since it is usually difficult to capture an all-in-focus image of a 3D scene directly, various multi-focus image fusion methods are employed to generate it from several images focusing at different depths. However, the performance of existing methods is barely satisfactory and often degrades for areas near the focused/defocused boundary (FDB). In this paper, a boundary aware method using deep neural network is proposed to overcome this problem. (1) Aiming to acquire improved fusion images, a 2-channel deep network is proposed to better extract the relative defocus information of the two source images. (2) After analyzing the different situations for patches far away from and near the FDB, we use two networks to handle them respectively. (3) To simulate the reality more precisely, a new approach of dataset generation is designed. Experiments demonstrate that the proposed method outperforms the state-of-the-art methods, both qualitatively and quantitatively.
Training end-to-end deep robot policies requires a lot of domain-, task-, and hardware-specific data, which is often costly to provide. In this work, we propose to tackle this issue by employing a deep neural network with a modular architecture, consisting of separate perception, policy, and trajectory parts. Each part of the system is trained fully on synthetic data or in simulation. The data is exchanged between parts of the system as low-dimensional latent representations of affordances and trajectories. The performance is then evaluated in a zero-shot transfer scenario using Franka Panda robot arm. Results demonstrate that a low-dimensional representation of scene affordances extracted from an RGB image is sufficient to successfully train manipulator policies. We also introduce a method for affordance dataset generation, which is easily generalizable to new tasks, objects and environments, and requires no manual pixel labeling.
An unsupervised shape analysis is proposed to learn concepts reflecting shape commonalities. Our approach is two-fold: i) a spatial topology analysis of point cloud segment constellations within objects is used in which constellations are decomposed and described in a hierarchical and symbolic manner. ii) A topology analysis of the description space is used in which segment decompositions are exposed in. Inspired by Persistent Homology, groups of shape commonality are revealed. Experiments show that extracted persistent commonality groups can feature semantically meaningful shape concepts; the generalization of the proposed approach is evaluated by different real-world datasets. We extend this by not only learning shape concepts using real-world data, but by also using mental simulation of artificial abstract objects for training purposes. This extended approach is unsupervised in two respects: label-agnostic (no label information is used) and instance-agnostic (no instances preselected by human supervision are used for training). Experiments show that concepts generated with mental simulation, generalize and discriminate real object observations. Consequently, a robot may train and learn its own internal representation of concepts regarding shape appearance in a self-driven and machine-centric manner while omitting the tedious process of supervised dataset generation including the ambiguity in instance labeling and selection.
The ability to segment unknown objects in depth images has potential to enhance robot skills in grasping and object tracking. Recent computer vision research has demonstrated that Mask R-CNN can be trained to segment specific categories of objects in RGB images when massive hand-labeled datasets are available. As generating these datasets is time consuming, we instead train with synthetic depth images. Many robots now use depth sensors, and recent results suggest training on synthetic depth data can transfer successfully to the real world. We present a method for automated dataset generation and rapidly generate a synthetic training dataset of 50,000 depth images and 320,000 object masks using simulated heaps of 3D CAD models. We train a variant of Mask R-CNN with domain randomization on the generated dataset to perform category-agnostic instance segmentation without any hand-labeled data and we evaluate the trained network, which we refer to as Synthetic Depth (SD) Mask R-CNN, on a set of real, high-resolution depth images of challenging, densely-cluttered bins containing objects with highly-varied geometry. SD Mask R-CNN outperforms point cloud clustering baselines by an absolute 15% in Average Precision and 20% in Average Recall on COCO benchmarks, and achieves performance levels similar to a Mask R-CNN trained on a massive, hand-labeled RGB dataset and fine-tuned on real images from the experimental setup. We deploy the model in an instance-specific grasping pipeline to demonstrate its usefulness in a robotics application. Code, the synthetic training dataset, and supplementary material are available at https://bit.ly/2letCuE.
In this paper, we propose a recursive framework to recognize facial expressions from images in real scenes. Unlike traditional approaches that typically focus on developing and refining algorithms for improving recognition performance on an existing dataset, we integrate three important components in a recursive manner: facial dataset generation, facial expression recognition model building, and interactive interfaces for testing and new data collection. To start with, we first create a candid-images-for-facial-expression (CIFE) dataset. We then apply a convolutional neural network (CNN) to CIFE and build a CNN model for web image expression classification. In order to increase the expression recognition accuracy, we also fine-tune the CNN model and thus obtain a better CNN facial expression recognition model. Based on the fine-tuned CNN model, we design a facial expression game engine and collect a new and more balanced dataset, GaMo. The images of this dataset are collected from the different expressions our game users make when playing the game. Finally, we evaluate the GaMo and CIFE datasets and show that our recursive framework can help build a better facial expression model for dealing with real scene facial expression tasks.
We present a dual contribution to the task of machine reading-comprehension: a technique for creating large-sized machine-comprehension (MC) datasets using paragraph-vector models; and a novel, hybrid neural-network architecture that combines the representation power of recurrent neural networks with the discriminative power of fully-connected multi-layered networks. We use the MC-dataset generation technique to build a dataset of around 2 million examples, for which we empirically determine the high-ceiling of human performance (around 91% accuracy), as well as the performance of a variety of computer models. Among all the models we have experimented with, our hybrid neural-network architecture achieves the highest performance (83.2% accuracy). The remaining gap to the human-performance ceiling provides enough room for future model improvements.
Selective weeding is one of the key challenges in the field of agriculture robotics. To accomplish this task, a farm robot should be able to accurately detect plants and to distinguish them between crop and weeds. Most of the promising state-of-the-art approaches make use of appearance-based models trained on large annotated datasets. Unfortunately, creating large agricultural datasets with pixel-level annotations is an extremely time consuming task, actually penalizing the usage of data-driven techniques. In this paper, we face this problem by proposing a novel and effective approach that aims to dramatically minimize the human intervention needed to train the detection and classification algorithms. The idea is to procedurally generate large synthetic training datasets randomizing the key features of the target environment (i.e., crop and weed species, type of soil, light conditions). More specifically, by tuning these model parameters, and exploiting a few real-world textures, it is possible to render a large amount of realistic views of an artificial agricultural scenario with no effort. The generated data can be directly used to train the model or to supplement real-world images. We validate the proposed methodology by using as testbed a modern deep learning based image segmentation architecture. We compare the classification results obtained using both real and synthetic images as training data. The reported results confirm the effectiveness and the potentiality of our approach.
This master thesis addresses the subject of automatically generating a dataset for image recognition, which takes a lot of time when being done manually. As the thesis was written with motivation from the context of the biodiversity workgroup at the City University of Applied Sciences Bremen, the classification of taxonomic entries was chosen as an exemplary use case. In order to automate the dataset creation, a prototype was conceptualized and implemented after working out knowledge basics and analyzing requirements for it. It makes use of an pre-trained abstract artificial intelligence which is able to sort out images that do not contain the desired content. Subsequent to the implementation and the automated dataset creation resulting from it, an evaluation was performed. Other, manually collected datasets were compared to the one the prototype produced in means of specifications and accuracy. The results were more than satisfactory and showed that automatically generating a dataset for image recognition is not only possible, but also might be a decent alternative to spending time and money in doing this task manually. At the very end of this work, an idea of how to use the principle of employing abstract artificial intelligences for step-by-step classification of deeper taxonomic layers in a productive system is presented and discussed.