Feras Dayoub

On the performance of ConvNet features for place recognition

After the incredible success of deep learning in the computer vision domain, there has been much interest in applying Convolutional Network (ConvNet) features in robotic fields such as visual navigation and SLAM. Unfortunately, there are fundamental differences and challenges involved. Computer vision datasets are very different in character to robotic camera data, real-time performance is essential, and performance priorities can be different. This paper comprehensively evaluates and compares the utility of three state-of-the-art ConvNets on the problems of particular relevance to navigation for robots; viewpoint-invariance and condition-invariance, and for the first time enables real-time place recognition performance using ConvNets with large maps by integrating a variety of existing (locality-sensitive hashing) and novel (semantic search space partitioning) optimization techniques. We present extensive experiments on four real world datasets cultivated to evaluate each of the specific challenges in place recognition. The results demonstrate that speed-ups of two orders of magnitude can be achieved with minimal accuracy degradation, enabling real-time performance. We confirm that networks trained for semantic place categorization also perform better at (specific) place recognition when faced with severe appearance changes and provide a reference for which networks and layers are optimal for different aspects of the place recognition problem.

Place Recognition with ConvNet Landmarks: Viewpoint-Robust, Condition-Robust, Training-Free

Place recognition has long been an incompletely solved problem in that all approaches involve significant compromises. Current methods address many but never all of the critical challenges of place recognition – viewpoint-invariance, condition-invariance and minimizing training requirements. Here we present an approach that adapts state-of-the-art object proposal techniques to identify potential landmarks within an image for place recognition. We use the astonishing power of convolutional neural network features to identify matching landmark proposals between images to perform place recognition over extreme appearance and viewpoint variations. Our system does not require any form of training, all components are generic enough to be used off-the-shelf. We present a range of challenging experiments in varied viewpoint and environmental conditions. We demonstrate superior performance to current state-of-the- art techniques. Furthermore, by building on existing and widely used recognition frameworks, this approach provides a highly compatible place recognition system with the potential for easy integration of other techniques such as object detection and semantic scene interpretation.

An adaptive appearance-based map for long-term topological localization of mobile robots

This work considers a mobile service robot which uses an appearance-based representation of its workplace as a map, where the current view and the map are used to estimate the current position in the environment. Due to the nature of real-world environments such as houses and offices, where the appearance keeps changing, the internal representation may become out of date after some time. To solve this problem the robot needs to be able to adapt its internal representation continually to the changes in the environment. This paper presents a method for creating an adaptive map for long-term appearance-based localization of a mobile robot using long-term and short-term memory concepts, with omni-directional vision as the external sensor.

DeepFruits: A Fruit Detection System Using Deep Neural Networks

This paper presents a novel approach to fruit detection using deep convolutional neural networks. The aim is to build an accurate, fast and reliable fruit detection system, which is a vital element of an autonomous agricultural robotic platform; it is a key element for fruit yield estimation and automated harvesting. Recent work in deep neural networks has led to the development of a state-of-the-art object detector termed Faster Region-based CNN (Faster R-CNN). We adapt this model, through transfer learning, for the task of fruit detection using imagery obtained from two modalities: colour (RGB) and Near-Infrared (NIR). Early and late fusion methods are explored for combining the multi-modal (RGB and NIR) information. This leads to a novel multi-modal Faster R-CNN model, which achieves state-of-the-art results compared to prior work with the F1 score, which takes into account both precision and recall performances improving from 0.807 to 0.838 for the detection of sweet pepper. In addition to improved accuracy, this approach is also much quicker to deploy for new fruits, as it requires bounding box annotation rather than pixel-level annotation (annotating bounding boxes is approximately an order of magnitude quicker to perform). The model is retrained to perform the detection of seven fruits, with the entire process taking four hours to annotate and train the new model per fruit.

Long-term experiments with an adaptive spherical view representation for navigation in changing environments

Real-world environments such as houses and offices change over time, meaning that a mobile robot’s map will become out of date. In previous work we introduced a method to update the reference views in a topological map so that a mobile robot could continue to localize itself in a changing environment using omni-directional vision. In this work we extend this longterm updating mechanism to incorporate a spherical metric representation of the observed visual features for each node in the topological map. Using multi-view geometry we are then able to estimate the heading of the robot, in order to enable navigation between the nodes of the map, and to simultaneously adapt the spherical view representation in response to environmental changes. The results demonstrate the persistent performance of the proposed system in a long-term experiment.

Evaluation of Features for Leaf Classification in Challenging Conditions

Fine-grained leaf classification has concentrated on the use of traditional shape and statistical features to classify ideal images. In this paper we evaluate the effectiveness of traditional hand-crafted features and propose the use of deep convolutional neural network (Conv Net) features. We introduce a range of condition variations to explore the robustness of these features, including: translation, scaling, rotation, shading and occlusion. Evaluations on the Flavia dataset demonstrate that in ideal imaging conditions, combining traditional and Conv Net features yields state-of-the art performance with an average accuracy of 97.3%±0:6% compared to traditional features which obtain an average accuracy of 91.2%±1:6%. Further experiments show that this combined classification approach consistently outperforms the best set of traditional features by an average of 5.7% for all of the evaluated condition variations.

Place categorization and semantic mapping on a mobile robot

In this paper we focus on the challenging problem of place categorization and semantic mapping on a robot without environment-specific training. Motivated by their ongoing success in various visual recognition tasks, we build our system upon a state-of-the-art convolutional network. We overcome its closed-set limitations by complementing the network with a series of one-vs-all classifiers that can learn to recognize new semantic classes online. Prior domain knowledge is incorporated by embedding the classification system into a Bayesian filter framework that also ensures temporal coherence. We evaluate the classification accuracy of the system on a robot that maps a variety of places on our campus in real-time. We show how semantic information can boost robotic object detection performance and how the semantic map can be used to modulate the robots behaviour during navigation tasks. The system is made available to the community as a ROS module.

Peduncle Detection of Sweet Pepper for Autonomous Crop Harvesting—Combined Color and 3-D Information

This letter presents a three-dimensional (3-D) visual detection method for the challenging task of detecting peduncles of sweet peppers (Capsicum annuum) in the field. Cutting the peduncle cleanly is one of the most difficult stages of the harvesting process, where the peduncle is the part of the crop that attaches it to the main stem of the plant. Accurate peduncle detection in 3-D space is, therefore, a vital step in reliable autonomous harvesting of sweet peppers, as this can lead to precise cutting while avoiding damage to the surrounding plant. This letter makes use of both color and geometry information acquired from an RGB-D sensor and utilizes a supervised-learning approach for the peduncle detection task. The performance of the proposed method is demonstrated and evaluated by using qualitative and quantitative results [the area-under-the-curve (AUC) of the detection precision-recall curve]. We are able to achieve an AUC of 0.71 for peduncle detection on field-grown sweet peppers. We release a set of manually annotated 3-D sweet pepper and peduncle images to assist the research community in performing further research on this topic.

Robotic detection and tracking of Crown-of-Thorns starfish

This paper presents a novel vision-based underwater robotic system for the identification and control of Crown- Of-Thorns starfish (COTS) in coral reef environments. COTS have been identified as one of the most significant threats to Australias Great Barrier Reef. These starfish literally eat coral, impacting large areas of reef and the marine ecosystem that depends on it. Evidence has suggested that land-based nutrient runoff has accelerated recent outbreaks of COTS requiring extensive use of divers to manually inject biological agents into the starfish in an attempt to control population numbers. Facilitating this control program using robotics is the goal of our research. In this paper we introduce a vision-based COTS detection and tracking system based on a Random Forest Classifier (RFC) trained on images from underwater footage. To track COTS with a moving camera, we embed the RFC in a particle filter detector and tracker where the predicted class probability of the RFC is used as an observation probability to weight the particles, and we use a sparse optical flow estimation for the prediction step of the filter. The system is experimentally evaluated in a realistic laboratory setup using a robotic arm that moves a camera at different speeds and heights over a range of real-size images of COTS in a reef environment.

A rapidly deployable classification system using visual data for the application of precision weed management

In this work we demonstrate a rapidly deployable weed classification system that uses visual data to enable autonomous precision weeding without making prior assumptions about which weed species are present in a given field. Previous work in this area relies on having prior knowledge of the weed species present in the field. This assumption cannot always hold true for every field, and thus limits the use of weed classification systems based on this assumption. In this work, we obviate this assumption and introduce a rapidly deployable approach able to operate on any field without any weed species assumptions prior to deployment. We present a three stage pipeline for the implementation of our weed classification system consisting of initial field surveillance, offline processing and selective labelling, and automated precision weeding. The key characteristic of our approach is the combination of plant clustering and selective labelling which is what enables our system to operate without prior weed species knowledge. Testing using field data we are able to label 12.3 times fewer images than traditional full labelling whilst reducing classification accuracy by only 14%.