In state-of-the-art deep neural networks, both feature normalization and feature attention have become ubiquitous. % with significant performance improvement shown in a vast amount of tasks. They are usually studied as separate modules, however. In this paper, we propose a light-weight integration between the two schema and present Attentive Normalization (AN). Instead of learning a single affine transformation, AN learns a mixture of affine transformations and utilizes their weighted-sum as the final affine transformation applied to re-calibrate features in an instance-specific way. The weights are learned by leveraging channel-wise feature attention. In experiments, we test the proposed AN using four representative neural architectures in the ImageNet-1000 classification benchmark and the MS-COCO 2017 object detection and instance segmentation benchmark. AN obtains consistent performance improvement for different neural architectures in both benchmarks with absolute increase of top-1 accuracy in ImageNet-1000 between 0.5\% and 2.7\%, and absolute increase up to 1.8\% and 2.2\% for bounding box and mask AP in MS-COCO respectively. We observe that the proposed AN provides a strong alternative to the widely used Squeeze-and-Excitation (SE) module. The source codes are publicly available at https://github.com/iVMCL/AOGNet-v2 (the ImageNet Classification Repo) and https://github.com/iVMCL/AttentiveNorm\_Detection (the MS-COCO Detection and Segmentation Repo).
In order to function in unstructured environments, robots need the ability to recognize unseen novel objects. We take a step in this direction by tackling the problem of segmenting unseen object instances in tabletop environments. However, the type of large-scale real-world dataset required for this task typically does not exist for most robotic settings, which motivates the use of synthetic data. We propose a novel method that separately leverages synthetic RGB and synthetic depth for unseen object instance segmentation. Our method is comprised of two stages where the first stage operates only on depth to produce rough initial masks, and the second stage refines these masks with RGB. Surprisingly, our framework is able to learn from synthetic RGB-D data where the RGB is non-photorealistic. To train our method, we introduce a large-scale synthetic dataset of random objects on tabletops. We show that our method, trained on this dataset, can produce sharp and accurate masks, outperforming state-of-the-art methods on unseen object instance segmentation. We also show that our method can segment unseen objects for robot grasping. Code, models and video can be found at https://rse-lab.cs.washington.edu/projects/unseen-object-instance-segmentation/.
Object detection is a critical part of visual scene understanding. The representation of the object in the detection task has important implications on the efficiency and feasibility of annotation, robustness to occlusion, pose, lighting, and other visual sources of semantic uncertainty, and effectiveness in real-world applications (e.g., autonomous driving). Popular object representations include 2D and 3D bounding boxes, polygons, splines, pixels, and voxels. Each have their strengths and weakness. In this work, we propose a new representation of objects based on the bivariate normal distribution. This distribution-based representation has the benefit of robust detection of highly-overlapping objects and the potential for improved downstream tracking and instance segmentation tasks due to the statistical representation of object edges. We provide qualitative evaluation of this representation for the object detection task and quantitative evaluation of its use in a baseline algorithm for the instance segmentation task.
Weakly supervised instance labeling using only image-level labels, in lieu of expensive fine-grained pixel annotations, is crucial in several applications including medical image analysis. In contrast to conventional instance segmentation scenarios in computer vision, the problems that we consider are characterized by a small number of training images and non-local patterns that lead to the diagnosis. In this paper, we explore the use of multiple instance learning (MIL) to design an instance label generator under this weakly supervised setting. Motivated by the observation that an MIL model can handle bags of varying sizes, we propose to repurpose an MIL model originally trained for bag-level classification to produce reliable predictions for single instances, i.e., bags of size $1$. To this end, we introduce a novel regularization strategy based on virtual adversarial training for improving MIL training, and subsequently develop a knowledge distillation technique for repurposing the trained MIL model. Using empirical studies on colon cancer and breast cancer detection from histopathological images, we show that the proposed approach produces high-quality instance-level prediction and significantly outperforms state-of-the MIL methods.
Object Detection has been a significant topic in computer vision. As the continuous development of Deep Learning, many advanced academic and industrial outcomes are established on localising and classifying the target objects, such as instance segmentation, video tracking and robotic vision. As the core concept of Deep Learning, Deep Neural Networks (DNNs) and associated training are highly integrated with task-driven modelling, having great effects on accurate detection. The main focus of improving detection performance is proposing DNNs with extra layers and novel topological connections to extract the desired features from input data. However, training these models can be computationally expensive and laborious progress as the complicated model architecture and enormous parameters. Besides, the dataset is another reason causing this issue and low detection accuracy, because of insufficient data samples or difficult instances. To address these training difficulties, this thesis presents two different approaches to improve the detection performance in the relatively light-weight way. As the intrinsic feature of data-driven in deep learning, the first approach is "slot-based image augmentation" to enrich the dataset with extra foreground and background combinations. Instead of the commonly used image flipping method, the proposed system achieved similar mAP improvement with less extra images which decrease training time. This proposed augmentation system has extra flexibility adapting to various scenarios and the performance-driven analysis provides an alternative aspect of conducting image augmentation
Agricultural applications such as yield prediction, precision agriculture and automated harvesting need systems able to infer the crop state from low-cost sensing devices. Proximal sensing using affordable cameras combined with computer vision has seen a promising alternative, strengthened after the advent of convolutional neural networks (CNNs) as an alternative for challenging pattern recognition problems in natural images. Considering fruit growing monitoring and automation, a fundamental problem is the detection, segmentation and counting of individual fruits in orchards. Here we show that for wine grapes, a crop presenting large variability in shape, color, size and compactness, grape clusters can be successfully detected, segmented and tracked using state-of-the-art CNNs. In a test set containing 408 grape clusters from images taken on a trellis-system based vineyard, we have reached an F 1 -score up to 0.91 for instance segmentation, a fine separation of each cluster from other structures in the image that allows a more accurate assessment of fruit size and shape. We have also shown as clusters can be identified and tracked along video sequences recording orchard rows. We also present a public dataset containing grape clusters properly annotated in 300 images and a novel annotation methodology for segmentation of complex objects in natural images. The presented pipeline for annotation, training, evaluation and tracking of agricultural patterns in images can be replicated for different crops and production systems. It can be employed in the development of sensing components for several agricultural and environmental applications.
We introduce the first approach to solve the challenging problem of unsupervised 4D visual scene understanding for complex dynamic scenes with multiple interacting people from multi-view video. Our approach simultaneously estimates a detailed model that includes a per-pixel semantically and temporally coherent reconstruction, together with instance-level segmentation exploiting photo-consistency, semantic and motion information. We further leverage recent advances in 3D pose estimation to constrain the joint semantic instance segmentation and 4D temporally coherent reconstruction. This enables per person semantic instance segmentation of multiple interacting people in complex dynamic scenes. Extensive evaluation of the joint visual scene understanding framework against state-of-the-art methods on challenging indoor and outdoor sequences demonstrates a significant (approx 40%) improvement in semantic segmentation, reconstruction and scene flow accuracy.
Most existing methods handle cell instance segmentation problems directly without relying on additional detection boxes. These methods generally fails to separate touching cells due to the lack of global understanding of the objects. In contrast, box-based instance segmentation solves this problem by combining object detection with segmentation. However, existing methods typically utilize anchor box-based detectors, which would lead to inferior instance segmentation performance due to the class imbalance issue. In this paper, we propose a new box-based cell instance segmentation method. In particular, we first detect the five pre-defined points of a cell via keypoints detection. Then we group these points according to a keypoint graph and subsequently extract the bounding box for each cell. Finally, cell segmentation is performed on feature maps within the bounding boxes. We validate our method on two cell datasets with distinct object shapes, and empirically demonstrate the superiority of our method compared to other instance segmentation techniques. Code is available at: https://github.com/yijingru/KG_Instance_Segmentation.
We present DetectFusion, an RGB-D SLAM system that runs in real-time and can robustly handle semantically known and unknown objects that can move dynamically in the scene. Our system detects, segments and assigns semantic class labels to known objects in the scene, while tracking and reconstructing them even when they move independently in front of the monocular camera. In contrast to related work, we achieve real-time computational performance on semantic instance segmentation with a novel method combining 2D object detection and 3D geometric segmentation. In addition, we propose a method for detecting and segmenting the motion of semantically unknown objects, thus further improving the accuracy of camera tracking and map reconstruction. We show that our method performs on par or better than previous work in terms of localization and object reconstruction accuracy, while achieving about 20 FPS even if the objects are segmented in each frame.
Object detection is a trendy branch of computer vision, especially on human recognition and pedestrian detection. Recognizing the complete body of a person has always been a difficult problem. Over the years, researchers proposed various methods, and recently, Mask R-CNN has made a breakthrough for instance segmentation. Based on Faster R-CNN, Mask R-CNN has been able to generate a segmentation mask for each instance. We propose an application to extracts multiple persons from images and videos for pleasant life scenes to grouping happy moments of people such as family or friends and a community for QOL (Quality Of Life). We likewise propose a methodology to put extracted images of persons into the new background. This enables a user to make a pleasant collection of happy facial expressions and actions of his/her family and friends in his/her life. Mask R-CNN detects all types of object masks from images. Then our algorithm considers only the target person and extracts a person only without obstacles, such as dogs in front of the person, and the user also can select multiple persons as their expectations. Our algorithm is effective for both an image and a video irrespective of the length of it. Our algorithm does not add any overhead to Mask R-CNN, running at 5 fps. We show examples of yoga-person in an image and a dancer in a dance-video frame. We hope our simple and effective approach would serve as a baseline for replacing the image background and help ease future research.
The ability to detect objects regardless of image distortions or weather conditions is crucial for real-world applications of deep learning like autonomous driving. We here provide an easy-to-use benchmark to assess how object detection models perform when image quality degrades. The three resulting benchmark datasets, termed Pascal-C, Coco-C and Cityscapes-C, contain a large variety of image corruptions. We show that a range of standard object detection models suffer a severe performance loss on corrupted images (down to 30--60\% of the original performance). However, a simple data augmentation trick---stylizing the training images---leads to a substantial increase in robustness across corruption type, severity and dataset. We envision our comprehensive benchmark to track future progress towards building robust object detection models. Benchmark, code and data are publicly available.
Instance segmentation is a promising yet challenging topic in computer vision. Recent approaches such as Mask R-CNN typically divide this problem into two parts -- a detection component and a mask generation branch, and mostly focus on the improvement of the detection part. In this paper, we present an approach that extends Mask R-CNN with five novel optimization techniques for improving the mask generation branch and reducing the conflicts between the mask branch and the detection component in training. These five techniques are independent to each other and can be flexibly utilized in building various instance segmentation architectures for increasing the overall accuracy. We demonstrate the effectiveness of our approach with tests on the COCO dataset.
We formulate object segmentation in video as a graph partitioning problem in space and time, in which nodes are pixels and their relations form local neighborhoods. We claim that the strongest cluster in this pixel-level graph represents the salient object segmentation. We compute the main cluster using a novel and fast 3D filtering technique that finds the spectral clustering solution, namely the principal eigenvector of the graph's adjacency matrix, without building the matrix explicitly - which would be intractable. Our method is based on the power iteration for finding the principal eigenvector of a matrix, which we prove is equivalent to performing a specific set of 3D convolutions in the space-time feature volume. This allows us to avoid creating the matrix and have a fast parallel implementation on GPU. We show that our method is much faster than classical power iteration applied directly on the adjacency matrix. Different from other works, ours is dedicated to preserving object consistency in space and time at the level of pixels. For that, it requires powerful pixel-wise features at the frame level. This makes it perfectly suitable for incorporating the output of a backbone network or other methods and fast-improving over their solution without supervision. In experiments, we obtain consistent improvement, with the same set of hyper-parameters, over the top state of the art methods on DAVIS-2016 dataset, both in unsupervised and semi-supervised tasks. We also achieve top results on the well-known SegTrackv2 dataset.
A major obstacle in instance segmentation is that existing methods often need many per-pixel labels in order to be effective. These labels require large human effort and for certain applications, such labels are not readily available. To address this limitation, we propose a novel framework that can effectively train with image-level labels, which are significantly cheaper to acquire. For instance, one can do an internet search for the term "car" and obtain many images where a car is present with minimal effort. Our framework consists of two stages: (1) train a classifier to generate pseudo masks for the objects of interest; (2) train a fully supervised Mask R-CNN on these pseudo masks. Our two main contribution are proposing a pipeline that is simple to implement and is amenable to different segmentation methods; and achieves new state-of-the-art results for this problem setup. Our results are based on evaluating our method on PASCAL VOC 2012, a standard dataset for weakly supervised methods, where we demonstrate major performance gains compared to existing methods with respect to mean average precision.
We propose a theoretical framework that generalizes simple and fast algorithms for hierarchical agglomerative clustering to weighted graphs with both attractive and repulsive interactions between the nodes. This framework defines GASP, a Generalized Algorithm for Signed graph Partitioning, and allows us to explore many combinations of different linkage criteria and cannot-link constraints. We prove the equivalence of existing clustering methods to some of those combinations and introduce new algorithms for combinations that have not been studied before. We study both theoretical and empirical properties of these combinations and prove that some of these define an ultrametric on the graph. We conduct a systematic comparison of various instantiations of GASP on a large variety of both synthetic and existing signed clustering problems, in terms of accuracy but also efficiency and robustness to noise. Lastly, we show that some of the algorithms included in our framework, when combined with the predictions from a CNN model, result in a simple bottom-up instance segmentation pipeline. Going all the way from pixels to final segments with a simple procedure, we achieve state-of-the-art accuracy on the CREMI 2016 EM segmentation benchmark without requiring domain-specific superpixels.
Current state-of-the-art instance segmentation methods are not suited for real-time applications like autonomous driving, which require fast execution times at high accuracy. Although the currently dominant proposal-based methods have high accuracy, they are slow and generate masks at a fixed and low resolution. Proposal-free methods, by contrast, can generate masks at high resolution and are often faster, but fail to reach the same accuracy as the proposal-based methods. In this work we propose a new clustering loss function for proposal-free instance segmentation. The loss function pulls the spatial embeddings of pixels belonging to the same instance together and jointly learns an instance-specific clustering bandwidth, maximizing the intersection-over-union of the resulting instance mask. When combined with a fast architecture, the network can perform instance segmentation in real-time while maintaining a high accuracy. We evaluate our method on the challenging Cityscapes benchmark and achieve top results (5\% improvement over Mask R-CNN) at more than 10 fps on 2MP images. Code will be available at https://github.com/davyneven/SpatialEmbeddings .
In object detection, the intersection over union (IoU) threshold is frequently used to define positives/negatives. The threshold used to train a detector defines its \textit{quality}. While the commonly used threshold of 0.5 leads to noisy (low-quality) detections, detection performance frequently degrades for larger thresholds. This paradox of high-quality detection has two causes: 1) overfitting, due to vanishing positive samples for large thresholds, and 2) inference-time quality mismatch between detector and test hypotheses. A multi-stage object detection architecture, the Cascade R-CNN, composed of a sequence of detectors trained with increasing IoU thresholds, is proposed to address these problems. The detectors are trained sequentially, using the output of a detector as training set for the next. This resampling progressively improves hypotheses quality, guaranteeing a positive training set of equivalent size for all detectors and minimizing overfitting. The same cascade is applied at inference, to eliminate quality mismatches between hypotheses and detectors. An implementation of the Cascade R-CNN without bells or whistles achieves state-of-the-art performance on the COCO dataset, and significantly improves high-quality detection on generic and specific object detection datasets, including VOC, KITTI, CityPerson, and WiderFace. Finally, the Cascade R-CNN is generalized to instance segmentation, with nontrivial improvements over the Mask R-CNN. To facilitate future research, two implementations are made available at \url{https://github.com/zhaoweicai/cascade-rcnn} (Caffe) and \url{https://github.com/zhaoweicai/Detectron-Cascade-RCNN} (Detectron).
Crater ellipticity determination is a complex and time consuming task that so far has evaded successful automation. We train a state of the art computer vision algorithm to identify craters in Lunar digital elevation maps and retrieve their sizes and 2D shapes. The computational backbone of the model is MaskRCNN, an "instance segmentation" general framework that detects craters in an image while simultaneously producing a mask for each crater that traces its outer rim. Our post-processing pipeline then finds the closest fitting ellipse to these masks, allowing us to retrieve the crater ellipticities. Our model is able to correctly identify 87\% of known craters in the longitude range we hid from the network during training and validation (test set), while predicting thousands of additional craters not present in our training data. Manual validation of a subset of these "new" craters indicates that a majority of them are real, which we take as an indicator of the strength of our model in learning to identify craters, despite incomplete training data. The crater size, ellipticity, and depth distributions predicted by our model are consistent with human-generated results. The model allows us to perform a large scale search for differences in crater diameter and shape distributions between the lunar highlands and maria, and we exclude any such differences with a high statistical significance. The predicted test set catalogue and trained model are available here: https://github.com/malidib/Craters_MaskRCNN/.
We propose a novel method for instance label segmentation of dense 3D voxel grids. We target volumetric scene representations, which have been acquired with depth sensors or multi-view stereo methods and which have been processed with semantic 3D reconstruction or scene completion methods. The main task is to learn shape information about individual object instances in order to accurately separate them, including connected and incompletely scanned objects. We solve the 3D instance-labeling problem with a multi-task learning strategy. The first goal is to learn an abstract feature embedding, which groups voxels with the same instance label close to each other while separating clusters with different instance labels from each other. The second goal is to learn instance information by densely estimating directional information of the instance's center of mass for each voxel. This is particularly useful to find instance boundaries in the clustering post-processing step, as well as, for scoring the segmentation quality for the first goal. Both synthetic and real-world experiments demonstrate the viability and merits of our approach. In fact, it achieves state-of-the-art performance on the ScanNet 3D instance segmentation benchmark.
Occlusion-aware instance-sensitive segmentation is a complex task generally split into region-based segmentations, by approximating instances as their bounding box. We address the showcase scenario of dense homogeneous layouts in which this approximation does not hold. In this scenario, outlining unoccluded instances by decoding a deep encoder becomes difficult, due to the translation invariance of convolutional layers and the lack of complexity in the decoder. We therefore propose a multicameral design composed of subtask-specific lightweight decoder and encoder-decoder units, coupled in cascade to encourage subtask-specific feature reuse and enforce a learning path within the decoding process. Furthermore, the state-of-the-art datasets for occlusion-aware instance segmentation contain real images with few instances and occlusions mostly due to objects occluding the background, unlike dense object layouts. We thus also introduce a synthetic dataset of dense homogeneous object layouts, namely Mikado, which extensibly contains more instances and inter-instance occlusions per image than these public datasets. Our extensive experiments on Mikado and public datasets show that ordinal multiscale units within the decoding process prove more effective than state-of-the-art design patterns for capturing position-sensitive representations. We also show that Mikado is plausible with respect to real-world problems, in the sense that it enables the learning of performance-enhancing representations transferable to real images, while drastically reducing the need of hand-made annotations for finetuning. The proposed dataset will be made publicly available.