Vanilla models for object detection and instance segmentation suffer from the heavy bias toward detecting frequent objects in the long-tailed setting. Existing methods address this issue mostly during training, e.g., by re-sampling or re-weighting. In this paper, we investigate a largely overlooked approach -- post-processing calibration of confidence scores. We propose NorCal, Normalized Calibration for long-tailed object detection and instance segmentation, a simple and straightforward recipe that reweighs the predicted scores of each class by its training sample size. We show that separately handling the background class and normalizing the scores over classes for each proposal are keys to achieving superior performance. On the LVIS dataset, NorCal can effectively improve nearly all the baseline models not only on rare classes but also on common and frequent classes. Finally, we conduct extensive analysis and ablation studies to offer insights into various modeling choices and mechanisms of our approach. Our code is publicly available at https://github.com/tydpan/NorCal/.
Recently, Vision Transformer and its variants have shown great promise on various computer vision tasks. The ability of capturing short- and long-range visual dependencies through self-attention is arguably the main source for the success. But it also brings challenges due to quadratic computational overhead, especially for the high-resolution vision tasks (e.g., object detection). In this paper, we present focal self-attention, a new mechanism that incorporates both fine-grained local and coarse-grained global interactions. Using this new mechanism, each token attends the closest surrounding tokens at fine granularity but the tokens far away at coarse granularity, and thus can capture both short- and long-range visual dependencies efficiently and effectively. With focal self-attention, we propose a new variant of Vision Transformer models, called Focal Transformer, which achieves superior performance over the state-of-the-art vision Transformers on a range of public image classification and object detection benchmarks. In particular, our Focal Transformer models with a moderate size of 51.1M and a larger size of 89.8M achieve 83.5 and 83.8 Top-1 accuracy, respectively, on ImageNet classification at 224x224 resolution. Using Focal Transformers as the backbones, we obtain consistent and substantial improvements over the current state-of-the-art Swin Transformers for 6 different object detection methods trained with standard 1x and 3x schedules. Our largest Focal Transformer yields 58.7/58.9 box mAPs and 50.9/51.3 mask mAPs on COCO mini-val/test-dev, and 55.4 mIoU on ADE20K for semantic segmentation, creating new SoTA on three of the most challenging computer vision tasks.
Modern top-performing object detectors depend heavily on backbone networks, whose advances bring consistent performance gains through exploring more effective network structures. In this paper, we propose a novel and flexible backbone framework, namely CBNetV2, to construct high-performance detectors using existing open-sourced pre-trained backbones under the pre-training fine-tuning paradigm. In particular, CBNetV2 architecture groups multiple identical backbones, which are connected through composite connections. Specifically, it integrates the high- and low-level features of multiple backbone networks and gradually expands the receptive field to more efficiently perform object detection. We also propose a better training strategy with assistant supervision for CBNet-based detectors. Without additional pre-training of the composite backbone, CBNetV2 can be adapted to various backbones (CNN-based vs. Transformer-based) and head designs of most mainstream detectors (one-stage vs. two-stage, anchor-based vs. anchor-free-based). Experiments provide strong evidence that, compared with simply increasing the depth and width of the network, CBNetV2 introduces a more efficient, effective, and resource-friendly way to build high-performance backbone networks. Particularly, our Dual-Swin-L achieves 59.4% box AP and 51.6% mask AP on COCO test-dev under the single-model and single-scale testing protocol, which is significantly better than the state-of-the-art result (57.7% box AP and 50.2% mask AP) achieved by Swin-L, while the training schedule is reduced by 6$\times$. With multi-scale testing, we push the current best single model result to a new record of 60.1% box AP and 52.3% mask AP without using extra training data. Code is available at https://github.com/VDIGPKU/CBNetV2.
The speed-accuracy Pareto curve of object detection systems have advanced through a combination of better model architectures, training and inference methods. In this paper, we methodically evaluate a variety of these techniques to understand where most of the improvements in modern detection systems come from. We benchmark these improvements on the vanilla ResNet-FPN backbone with RetinaNet and RCNN detectors. The vanilla detectors are improved by 7.7% in accuracy while being 30% faster in speed. We further provide simple scaling strategies to generate family of models that form two Pareto curves, named RetinaNet-RS and Cascade RCNN-RS. These simple rescaled detectors explore the speed-accuracy trade-off between the one-stage RetinaNet detectors and two-stage RCNN detectors. Our largest Cascade RCNN-RS models achieve 52.9% AP with a ResNet152-FPN backbone and 53.6% with a SpineNet143L backbone. Finally, we show the ResNet architecture, with three minor architectural changes, outperforms EfficientNet as the backbone for object detection and instance segmentation systems.
Compared to many other dense prediction tasks, e.g., semantic segmentation, it is the arbitrary number of instances that has made instance segmentation much more challenging. In order to predict a mask for each instance, mainstream approaches either follow the 'detect-then-segment' strategy (e.g., Mask R-CNN), or predict embedding vectors first then cluster pixels into individual instances. In this paper, we view the task of instance segmentation from a completely new perspective by introducing the notion of "instance categories", which assigns categories to each pixel within an instance according to the instance's location. With this notion, we propose segmenting objects by locations (SOLO), a simple, direct, and fast framework for instance segmentation with strong performance. We derive a few SOLO variants (e.g., Vanilla SOLO, Decoupled SOLO, Dynamic SOLO) following the basic principle. Our method directly maps a raw input image to the desired object categories and instance masks, eliminating the need for the grouping post-processing or the bounding box detection. Our approach achieves state-of-the-art results for instance segmentation in terms of both speed and accuracy, while being considerably simpler than the existing methods. Besides instance segmentation, our method yields state-of-the-art results in object detection (from our mask byproduct) and panoptic segmentation. We further demonstrate the flexibility and high-quality segmentation of SOLO by extending it to perform one-stage instance-level image matting. Code is available at: https://git.io/AdelaiDet
Instance segmentation of images is an important tool for automated scene understanding. Neural networks are usually trained to optimize their overall performance in terms of accuracy. Meanwhile, in applications such as automated driving, an overlooked pedestrian seems more harmful than a falsely detected one. In this work, we present a false negative detection method for image sequences based on inconsistencies in time series of tracked instances given the availability of image sequences in online applications. As the number of instances can be greatly increased by this algorithm, we apply a false positive pruning using uncertainty estimates aggregated over instances. To this end, instance-wise metrics are constructed which characterize uncertainty and geometry of a given instance or are predicated on depth estimation. The proposed method serves as a post-processing step applicable to any neural network that can also be trained on single frames only. In our tests, we obtain an improved trade-off between false negative and false positive instances by our fused detection approach in comparison to the use of an ordinary score value provided by the instance segmentation network during inference.
Indoor panorama typically consists of human-made structures parallel or perpendicular to gravity. We leverage this phenomenon to approximate the scene in a 360-degree image with (H)orizontal-planes and (V)ertical-planes. To this end, we propose an effective divide-and-conquer strategy that divides pixels based on their plane orientation estimation; then, the succeeding instance segmentation module conquers the task of planes clustering more easily in each plane orientation group. Besides, parameters of V-planes depend on camera yaw rotation, but translation-invariant CNNs are less aware of the yaw change. We thus propose a yaw-invariant V-planar reparameterization for CNNs to learn. We create a benchmark for indoor panorama planar reconstruction by extending existing 360 depth datasets with ground truth H\&V-planes (referred to as PanoH&V dataset) and adopt state-of-the-art planar reconstruction methods to predict H\&V-planes as our baselines. Our method outperforms the baselines by a large margin on the proposed dataset.
In this work we propose a novel and fully automated method for extracting the yarn geometrical features in woven composites so that a direct parametrization of the textile reinforcement is achieved (e.g., FE mesh). Thus, our aim is not only to perform yarn segmentation from tomographic images but rather to provide a complete descriptive modeling of the fabric. As such, this direct approach improves on previous methods that use voxel-wise masks as intermediate representations followed by re-meshing operations (yarn envelope estimation). The proposed approach employs two deep neural network architectures (U-Net and Mask RCNN). First, we train the U-Net to generate synthetic CT images from the corresponding FE simulations. This allows to generate large quantities of annotated data without requiring costly manual annotations. This data is then used to train the Mask R-CNN, which is focused on predicting contour points around each of the yarns in the image. Experimental results show that our method is accurate and robust for performing yarn instance segmentation on CT images, this is further validated by quantitative and qualitative analyses.
Semantic, instance, and panoptic segmentations have been addressed using different and specialized frameworks despite their underlying connections. This paper presents a unified, simple, and effective framework for these essentially similar tasks. The framework, named K-Net, segments both instances and semantic categories consistently by a group of learnable kernels, where each kernel is responsible for generating a mask for either a potential instance or a stuff class. To remedy the difficulties of distinguishing various instances, we propose a kernel update strategy that enables each kernel dynamic and conditional on its meaningful group in the input image. K-Net can be trained in an end-to-end manner with bipartite matching, and its training and inference are naturally NMS-free and box-free. Without bells and whistles, K-Net surpasses all previous published state-of-the-art single-model results of panoptic segmentation on MS COCO test-dev split and semantic segmentation on ADE20K val split with 55.2% PQ and 54.3% mIoU, respectively. Its instance segmentation performance is also on par with Cascade Mask R-CNN on MS COCO with 60%-90% faster inference speeds. Code and models will be released at https://github.com/ZwwWayne/K-Net/.
Arbitrary-shaped text detection is a challenging task since curved texts in the wild are of the complex geometric layouts. Existing mainstream methods follow the instance segmentation pipeline to obtain the text regions. However, arbitraryshaped texts are difficult to be depicted through one single segmentation network because of the varying scales. In this paper, we propose a two-stage segmentation-based detector, termed as NASK (Need A Second looK), for arbitrary-shaped text detection. Compared to the traditional single-stage segmentation network, our NASK conducts the detection in a coarse-to-fine manner with the first stage segmentation spotting the rectangle text proposals and the second one retrieving compact representations. Specifically, NASK is composed of a Text Instance Segmentation (TIS) network (1st stage), a Geometry-aware Text RoI Alignment (GeoAlign) module, and a Fiducial pOint eXpression (FOX) module (2nd stage). Firstly, TIS extracts the augmented features with a novel Group Spatial and Channel Attention (GSCA) module and conducts instance segmentation to obtain rectangle proposals. Then, GeoAlign converts these rectangles into the fixed size and encodes RoI-wise feature representation. Finally, FOX disintegrates the text instance into serval pivotal geometrical attributes to refine the detection results. Extensive experimental results on three public benchmarks including Total-Text, SCUTCTW1500, and ICDAR 2015 verify that our NASK outperforms recent state-of-the-art methods.
Recently, the vision transformer has achieved great success by pushing the state-of-the-art of various vision tasks. One of the most challenging problems in the vision transformer is that the large sequence length of image tokens leads to high computational cost (quadratic complexity). A popular solution to this problem is to use a single pooling operation to reduce the sequence length. This paper considers how to improve existing vision transformers, where the pooled feature extracted by a single pooling operation seems less powerful. To this end, we note that pyramid pooling has been demonstrated to be effective in various vision tasks owing to its powerful ability in context abstraction. However, pyramid pooling has not been explored in backbone network design. To bridge this gap, we propose to adapt pyramid pooling to Multi-Head Self-Attention (MHSA) in the vision transformer, simultaneously reducing the sequence length and capturing powerful contextual features. Plugged with our pooling-based MHSA, we build a universal vision transformer backbone, dubbed Pyramid Pooling Transformer (P2T). Extensive experiments demonstrate that, when applied P2T as the backbone network, it shows substantial superiority in various vision tasks such as image classification, semantic segmentation, object detection, and instance segmentation, compared to previous CNN- and transformer-based networks. The code will be released at https://github.com/yuhuan-wu/P2T.
Although instance segmentation has made considerable advancement over recent years, it's still a challenge to design high accuracy algorithms with real-time performance. In this paper, we propose a real-time instance segmentation framework termed OrienMask. Upon the one-stage object detector YOLOv3, a mask head is added to predict some discriminative orientation maps, which are explicitly defined as spatial offset vectors for both foreground and background pixels. Thanks to the discrimination ability of orientation maps, masks can be recovered without the need for extra foreground segmentation. All instances that match with the same anchor size share a common orientation map. This special sharing strategy reduces the amortized memory utilization for mask predictions but without loss of mask granularity. Given the surviving box predictions after NMS, instance masks can be concurrently constructed from the corresponding orientation maps with low complexity. Owing to the concise design for mask representation and its effective integration with the anchor-based object detector, our method is qualified under real-time conditions while maintaining competitive accuracy. Experiments on COCO benchmark show that OrienMask achieves 34.8 mask AP at the speed of 42.7 fps evaluated with a single RTX 2080 Ti. The code is available at https://github.com/duwt/OrienMask.
Recent advances in bioimaging have provided scientists a superior high spatial-temporal resolution to observe dynamics of living cells as 3D volumetric videos. Unfortunately, the 3D biomedical video analysis is lagging, impeded by resource insensitive human curation using off-the-shelf 3D analytic tools. Herein, biologists often need to discard a considerable amount of rich 3D spatial information by compromising on 2D analysis via maximum intensity projection. Recently, pixel embedding-based cell instance segmentation and tracking provided a neat and generalizable computing paradigm for understanding cellular dynamics. In this work, we propose a novel spatial-temporal voxel-embedding (VoxelEmbed) based learning method to perform simultaneous cell instance segmenting and tracking on 3D volumetric video sequences. Our contribution is in four-fold: (1) The proposed voxel embedding generalizes the pixel embedding with 3D context information; (2) Present a simple multi-stream learning approach that allows effective spatial-temporal embedding; (3) Accomplished an end-to-end framework for one-stage 3D cell instance segmentation and tracking without heavy parameter tuning; (4) The proposed 3D quantification is memory efficient via a single GPU with 12 GB memory. We evaluate our VoxelEmbed method on four 3D datasets (with different cell types) from the ISBI Cell Tracking Challenge. The proposed VoxelEmbed method achieved consistent superior overall performance (OP) on two densely annotated datasets. The performance is also competitive on two sparsely annotated cohorts with 20.6% and 2% of data-set having segmentation annotations. The results demonstrate that the VoxelEmbed method is a generalizable and memory-efficient solution.
Recently, query based deep networks catch lots of attention owing to their end-to-end pipeline and competitive results on several fundamental computer vision tasks, such as object detection, semantic segmentation, and instance segmentation. However, how to establish a query based video instance segmentation (VIS) framework with elegant architecture and strong performance remains to be settled. In this paper, we present \textbf{QueryTrack} (i.e., tracking instances as queries), a unified query based VIS framework fully leveraging the intrinsic one-to-one correspondence between instances and queries in QueryInst. The proposed method obtains 52.7 / 52.3 AP on YouTube-VIS-2019 / 2021 datasets, which wins the 2-nd place in the YouTube-VIS Challenge at CVPR 2021 \textbf{with a single online end-to-end model, single scale testing \& modest amount of training data}. We also provide QueryTrack-ResNet-50 baseline results on YouTube-VIS-2021 val set as references for the VIS community.
In this work we present a novel solution for Video Instance Segmentation(VIS), that is automatically generating instance level segmentation masks along with object class and tracking them in a video. Our method improves the masks from segmentation and propagation branches in an online manner using the Mask Selection Network (MSN) hence limiting the noise accumulation during mask tracking. We propose an effective design of MSN by using patch-based convolutional neural network. The network is able to distinguish between very subtle differences between the masks and choose the better masks out of the associated masks accurately. Further, we make use of temporal consistency and process the video sequences in both forward and reverse manner as a post processing step to recover lost objects. The proposed method can be used to adapt any video object segmentation method for the task of VIS. Our method achieves a score of 49.1 mAP on 2021 YouTube-VIS Challenge and was ranked third place among more than 30 global teams. Our code will be available at https://github.com/SHI-Labs/Mask-Selection-Networks.
Aiming at facilitating a real-world, ever-evolving and scalable autonomous driving system, we present a large-scale dataset for standardizing the evaluation of different self-supervised and semi-supervised approaches by learning from raw data, which is the first and largest dataset to date. Existing autonomous driving systems heavily rely on `perfect' visual perception models (i.e., detection) trained using extensive annotated data to ensure safety. However, it is unrealistic to elaborately label instances of all scenarios and circumstances (i.e., night, extreme weather, cities) when deploying a robust autonomous driving system. Motivated by recent advances of self-supervised and semi-supervised learning, a promising direction is to learn a robust detection model by collaboratively exploiting large-scale unlabeled data and few labeled data. Existing datasets either provide only a small amount of data or covers limited domains with full annotation, hindering the exploration of large-scale pre-trained models. Here, we release a Large-Scale 2D Self/semi-supervised Object Detection dataset for Autonomous driving, named as SODA10M, containing 10 million unlabeled images and 20K images labeled with 6 representative object categories. To improve diversity, the images are collected within 27833 driving hours under different weather conditions, periods and location scenes of 32 different cities. We provide extensive experiments and deep analyses of existing popular self/semi-supervised approaches, and give some interesting findings in autonomous driving scope. Experiments show that SODA10M can serve as a promising pre-training dataset for different self-supervised learning methods, which gives superior performance when fine-tuning with different downstream tasks (i.e., detection, semantic/instance segmentation) in autonomous driving domain. More information can refer to https://soda-2d.github.io.
Boundary-based instance segmentation has drawn much attention since of its attractive efficiency. However, existing methods suffer from the difficulty in long-distance regression. In this paper, we propose a coarse-to-fine module to address the problem. Approximate boundary points are generated at the coarse stage and then features of these points are sampled and fed to a refined regressor for fine prediction. It is end-to-end trainable since differential sampling operation is well supported in the module. Furthermore, we design a holistic boundary-aware branch and introduce instance-agnostic supervision to assist regression. Equipped with ResNet-101, our approach achieves 31.7\% mask AP on COCO dataset with single-scale training and testing, outperforming the baseline 1.3\% mask AP with less than 1\% additional parameters and GFLOPs. Experiments also show that our proposed method achieves competitive performance compared to existing boundary-based methods with a lightweight design and a simple pipeline.
Following their success in natural language processing, transformers have recently shown much promise for computer vision. The self-attention operation underlying transformers yields global interactions between all tokens ,i.e. words or image patches, and enables flexible modelling of image data beyond the local interactions of convolutions. This flexibility, however, comes with a quadratic complexity in time and memory, hindering application to long sequences and high-resolution images. We propose a "transposed" version of self-attention that operates across feature channels rather than tokens, where the interactions are based on the cross-covariance matrix between keys and queries. The resulting cross-covariance attention (XCA) has linear complexity in the number of tokens, and allows efficient processing of high-resolution images. Our cross-covariance image transformer (XCiT) is built upon XCA. It combines the accuracy of conventional transformers with the scalability of convolutional architectures. We validate the effectiveness and generality of XCiT by reporting excellent results on multiple vision benchmarks, including image classification and self-supervised feature learning on ImageNet-1k, object detection and instance segmentation on COCO, and semantic segmentation on ADE20k.
Challenges have become the state-of-the-art approach to benchmark image analysis algorithms in a comparative manner. While the validation on identical data sets was a great step forward, results analysis is often restricted to pure ranking tables, leaving relevant questions unanswered. Specifically, little effort has been put into the systematic investigation on what characterizes images in which state-of-the-art algorithms fail. To address this gap in the literature, we (1) present a statistical framework for learning from challenges and (2) instantiate it for the specific task of instrument instance segmentation in laparoscopic videos. Our framework relies on the semantic meta data annotation of images, which serves as foundation for a General Linear Mixed Models (GLMM) analysis. Based on 51,542 meta data annotations performed on 2,728 images, we applied our approach to the results of the Robust Medical Instrument Segmentation Challenge (ROBUST-MIS) challenge 2019 and revealed underexposure, motion and occlusion of instruments as well as the presence of smoke or other objects in the background as major sources of algorithm failure. Our subsequent method development, tailored to the specific remaining issues, yielded a deep learning model with state-of-the-art overall performance and specific strengths in the processing of images in which previous methods tended to fail. Due to the objectivity and generic applicability of our approach, it could become a valuable tool for validation in the field of medical image analysis and beyond. and segmentation of small, crossing, moving and transparent instrument(s) (parts).
Attention mechanism of late has been quite popular in the computer vision community. A lot of work has been done to improve the performance of the network, although almost always it results in increased computational complexity. In this paper, we propose a new attention module that not only achieves the best performance but also has lesser parameters compared to most existing models. Our attention module can easily be integrated with other convolutional neural networks because of its lightweight nature. The proposed network named Dual Multi Scale Attention Network (DMSANet) is comprised of two parts: the first part is used to extract features at various scales and aggregate them, the second part uses spatial and channel attention modules in parallel to adaptively integrate local features with their global dependencies. We benchmark our network performance for Image Classification on ImageNet dataset, Object Detection and Instance Segmentation both on MS COCO dataset.