Mask R-CNN has recently achieved great success in the field of instance segmentation. However, weaknesses of the algorithm have been repeatedly pointed out as well, especially in the segmentation of long, sparse objects whose orientation is not exclusively horizontal or vertical. We present here an approach that significantly improves the performance of the algorithm by first pre-segmenting the images with a PSPNet algorithm. To further improve its prediction, we have developed our own cost functions and heuristics in the form of training strategies, which can prevent so-called (early) overfitting and achieve a more targeted convergence. Furthermore, due to the high variance of the images, especially for PSPNet, we aimed to develop strategies for a high robustness and generalization, which are also presented here.
In this paper, we bring forward the use of the recently developed Signature Transform as a way to measure the similarity between image distributions and provide detailed acquaintance and extensive evaluations. We are the first to pioneer RMSE and MAE Signature, along with log-signature as an alternative to measure GAN convergence, a problem that has been extensively studied. We are also forerunners to introduce analytical measures based on statistics to study the goodness of fit of the GAN sample distribution that are both efficient and effective. Current GAN measures involve lots of computation normally done at the GPU and are very time consuming. In contrast, we diminish the computation time to the order of seconds and computation is done at the CPU achieving the same level of goodness. Lastly, a PCA adaptive t-SNE approach, which is novel in this context, is also proposed for data visualization.
Video instance segmentation is a challenging task that extends image instance segmentation to the video domain. Existing methods either rely only on single-frame information for the detection and segmentation subproblems or handle tracking as a separate post-processing step, which limit their capability to fully leverage and share useful spatial-temporal information for all the subproblems. In this paper, we propose a novel graph-neural-network (GNN) based method to handle the aforementioned limitation. Specifically, graph nodes representing instance features are used for detection and segmentation while graph edges representing instance relations are used for tracking. Both inter and intra-frame information is effectively propagated and shared via graph updates and all the subproblems (i.e. detection, segmentation and tracking) are jointly optimized in an unified framework. The performance of our method shows great improvement on the YoutubeVIS validation dataset compared to existing methods and achieves 35.2% AP with a ResNet-50 backbone, operating at 22 FPS. Code is available at http://github.com/lucaswithai/visgraph.git .
The study of multi-task learning has drawn great attention from the community. Despite the remarkable progress, the challenge of optimally learning different tasks simultaneously remains to be explored. Previous works attempt to modify the gradients from different tasks. Yet these methods give a subjective assumption of the relationship between tasks, and the modified gradient may be less accurate. In this paper, we introduce Stochastic Task Allocation~(STA), a mechanism that addresses this issue by a task allocation approach, in which each sample is randomly allocated a subset of tasks. For further progress, we propose Interleaved Stochastic Task Allocation~(ISTA) to iteratively allocate all tasks to each example during several consecutive iterations. We evaluate STA and ISTA on various datasets and applications: NYUv2, Cityscapes, and COCO for scene understanding and instance segmentation. Our experiments show both STA and ISTA outperform current state-of-the-art methods. The code will be available.
Top-down instance segmentation framework has shown its superiority in object detection compared to the bottom-up framework. While it is efficient in addressing over-segmentation, top-down instance segmentation suffers from over-crop problem. However, a complete segmentation mask is crucial for biological image analysis as it delivers important morphological properties such as shapes and volumes. In this paper, we propose a region proposal rectification (RPR) module to address this challenging incomplete segmentation problem. In particular, we offer a progressive ROIAlign module to introduce neighbor information into a series of ROIs gradually. The ROI features are fed into an attentive feed-forward network (FFN) for proposal box regression. With additional neighbor information, the proposed RPR module shows significant improvement in correction of region proposal locations and thereby exhibits favorable instance segmentation performances on three biological image datasets compared to state-of-the-art baseline methods. Experimental results demonstrate that the proposed RPR module is effective in both anchor-based and anchor-free top-down instance segmentation approaches, suggesting the proposed method can be applied to general top-down instance segmentation of biological images.
Industrial bin picking is a challenging task that requires accurate and robust segmentation of individual object instances. Particularly, industrial objects can have irregular shapes, that is, thin and concave, whereas in bin-picking scenarios, objects are often closely packed with strong occlusion. To address these challenges, we formulate a novel part-aware instance segmentation pipeline. The key idea is to decompose industrial objects into correlated approximate convex parts and enhance the object-level segmentation with part-level segmentation. We design a part-aware network to predict part masks and part-to-part offsets, followed by a part aggregation module to assemble the recognized parts into instances. To guide the network learning, we also propose an automatic label decoupling scheme to generate ground-truth part-level labels from instance-level labels. Finally, we contribute the first instance segmentation dataset, which contains a variety of industrial objects that are thin and have non-trivial shapes. Extensive experimental results on various industrial objects demonstrate that our method can achieve the best segmentation results compared with the state-of-the-art approaches.
Instance segmentation and classification of nuclei is an important task in computational pathology. We show that StarDist, a deep learning nuclei segmentation method originally developed for fluorescence microscopy, can be extended and successfully applied to histopathology images. This is substantiated by conducting experiments on the Lizard dataset, and through entering the Colon Nuclei Identification and Counting (CoNIC) challenge 2022, where our approach achieved the first spot on the leaderboard for the segmentation and classification task for both the preliminary and final test phase.
Nuclei segmentation and classification is the first and most crucial step that is utilized for many different microscopy medical analysis applications. However, it suffers from many issues such as the segmentation of small objects, imbalance, and fine-grained differences between types of nuclei. In this paper, multiple different contributions were done tackling these problems present. Firstly, the recently released "ConvNeXt" was used as the encoder for HoVer-Net model since it leverages the key components of transformers that make them perform well. Secondly, to enhance the visual differences between nuclei, a multi-channel color space-based approach is used to aid the model in extracting distinguishing features. Thirdly, Unified Focal loss (UFL) was used to tackle the background-foreground imbalance. Finally, Sharpness-Aware Minimization (SAM) was used to ensure generalizability of the model. Overall, we were able to outperform the current state-of-the-art (SOTA), HoVer-Net, on the preliminary test set of the CoNiC Challenge 2022 by 12.489% mPQ+.
Wood logs picking is a challenging task to automate. Indeed, logs usually come in cluttered configurations, randomly orientated and overlapping. Recent work on log picking automation usually assume that the logs' pose is known, with little consideration given to the actual perception problem. In this paper, we squarely address the latter, using a data-driven approach. First, we introduce a novel dataset, named TimberSeg 1.0, that is densely annotated, i.e., that includes both bounding boxes and pixel-level mask annotations for logs. This dataset comprises 220 images with 2500 individually segmented logs. Using our dataset, we then compare three neural network architectures on the task of individual logs detection and segmentation; two region-based methods and one attention-based method. Unsurprisingly, our results show that axis-aligned proposals, failing to take into account the directional nature of logs, underperform with 19.03 mAP. A rotation-aware proposal method significantly improve results to 31.83 mAP. More interestingly, a Transformer-based approach, without any inductive bias on rotations, outperformed the two others, achieving a mAP of 57.53 on our dataset. Our use case demonstrates the limitations of region-based approaches for cluttered, elongated objects. It also highlights the potential of attention-based methods on this specific task, as they work directly at the pixel-level. These encouraging results indicate that such a perception system could be used to assist the operators on the short-term, or to fully automate log picking operations in the future.
Video Instance Segmentation (VIS) aims to simultaneously classify, segment, and track multiple object instances in videos. Recent clip-level VIS takes a short video clip as input each time showing stronger performance than frame-level VIS (tracking-by-segmentation), as more temporal context from multiple frames is utilized. Yet, most clip-level methods are neither end-to-end learnable nor real-time. These limitations are addressed by the recent VIS transformer (VisTR) which performs VIS end-to-end within a clip. However, VisTR suffers from long training time due to its frame-wise dense attention. In addition, VisTR is not fully end-to-end learnable in multiple video clips as it requires a hand-crafted data association to link instance tracklets between successive clips. This paper proposes EfficientVIS, a fully end-to-end framework with efficient training and inference. At the core are tracklet query and tracklet proposal that associate and segment regions-of-interest (RoIs) across space and time by an iterative query-video interaction. We further propose a correspondence learning that makes tracklets linking between clips end-to-end learnable. Compared to VisTR, EfficientVIS requires 15x fewer training epochs while achieving state-of-the-art accuracy on the YouTube-VIS benchmark. Meanwhile, our method enables whole video instance segmentation in a single end-to-end pass without data association at all.
Existing state-of-the-art 3D instance segmentation methods perform semantic segmentation followed by grouping. The hard predictions are made when performing semantic segmentation such that each point is associated with a single class. However, the errors stemming from hard decision propagate into grouping that results in (1) low overlaps between the predicted instance with the ground truth and (2) substantial false positives. To address the aforementioned problems, this paper proposes a 3D instance segmentation method referred to as SoftGroup by performing bottom-up soft grouping followed by top-down refinement. SoftGroup allows each point to be associated with multiple classes to mitigate the problems stemming from semantic prediction errors and suppresses false positive instances by learning to categorize them as background. Experimental results on different datasets and multiple evaluation metrics demonstrate the efficacy of SoftGroup. Its performance surpasses the strongest prior method by a significant margin of +6.2% on the ScanNet v2 hidden test set and +6.8% on S3DIS Area 5 in terms of AP_50. SoftGroup is also fast, running at 345ms per scan with a single Titan X on ScanNet v2 dataset. The source code and trained models for both datasets are available at \url{https://github.com/thangvubk/SoftGroup.git}.
Cancer is one of the leading causes of death in the developed world. Cancer diagnosis is performed through the microscopic analysis of a sample of suspicious tissue. This process is time consuming and error prone, but Deep Learning models could be helpful for pathologists during cancer diagnosis. We propose to change the CenterNet2 object detection model to also perform instance segmentation, which we call SegCenterNet2. We train SegCenterNet2 in the CoNIC challenge dataset and show that it performs better than Mask R-CNN in the competition metrics.
We revisit the one- and two-stage detector distillation tasks and present a simple and efficient semantic-aware framework to fill the gap between them. We address the pixel-level imbalance problem by designing the category anchor to produce a representative pattern for each category and regularize the topological distance between pixels and category anchors to further tighten their semantic bonds. We name our method SEA (SEmantic-aware Alignment) distillation given the nature of abstracting dense fine-grained information by semantic reliance to well facilitate distillation efficacy. SEA is well adapted to either detection pipeline and achieves new state-of-the-art results on the challenging COCO object detection task on both one- and two-stage detectors. Its superior performance on instance segmentation further manifests the generalization ability. Both 2x-distilled RetinaNet and FCOS with ResNet50-FPN outperform their corresponding 3x ResNet101-FPN teacher, arriving 40.64 and 43.06 AP, respectively. Code will be made publicly available.
We address the problem of automated nuclear segmentation, classification, and quantification from Haematoxylin and Eosin stained histology images, which is of great relevance for several downstream computational pathology applications. In this work, we present a solution framed as a simultaneous semantic and instance segmentation framework. Our solution is part of the Colon Nuclei Identification and Counting (CoNIC) Challenge. We first train a semantic and instance segmentation model separately. Our framework uses as backbone HoverNet and Cascade Mask-RCNN models. We then ensemble the results with a custom Non-Maximum Suppression embedding (NMS). In our framework, the semantic model computes a class prediction for the cells whilst the instance model provides a refined segmentation. We demonstrate, through our experimental results, that our model outperforms the provided baselines by a large margin.
Automated cell nucleus segmentation and classification are required to assist pathologists in their decision making. The Colon Nuclei Identification and Counting Challenge 2022 (CoNIC Challenge 2022) supports the development and comparability of segmentation and classification methods for histopathological images. In this contribution, we describe our CoNIC Challenge 2022 method ciscNet to segment, classify and count cell nuclei, and report preliminary evaluation results. Our code is available at https://git.scc.kit.edu/ciscnet/ciscnet-conic-2022.
Weakly supervised instance segmentation has gained popularity because it reduces high annotation cost of pixel-level masks required for model training. Recent approaches for weakly supervised instance segmentation detect and segment objects using appearance information obtained from a static image. However, it poses the challenge of identifying objects with a non-discriminatory appearance. In this study, we address this problem by using motion information from image sequences. We propose a two-stream encoder that leverages appearance and motion features extracted from images and optical flows. Additionally, we propose a novel pairwise loss that considers both appearance and motion information to supervise segmentation. We conducted extensive evaluations on the YouTube-VIS 2019 benchmark dataset. Our results demonstrate that the proposed method improves the Average Precision of the state-of-the-art method by 3.1.
A standard hardware bottleneck when training deep neural networks is GPU memory. The bulk of memory is occupied by caching intermediate tensors for gradient computation in the backward pass. We propose a novel method to reduce this footprint - Dropping Intermediate Tensors (DropIT). DropIT drops min-k elements of the intermediate tensors and approximates gradients from the sparsified tensors in the backward pass. Theoretically, DropIT reduces noise on estimated gradients and therefore has a higher rate of convergence than vanilla-SGD. Experiments show that we can drop up to 90\% of the intermediate tensor elements in fully-connected and convolutional layers while achieving higher testing accuracy for Visual Transformers and Convolutional Neural Networks on various tasks (e.g., classification, object detection, instance segmentation). Our code and models are available at https://github.com/chenjoya/dropit.
Calibrated confidence estimates obtained from neural networks are crucial, particularly for safety-critical applications such as autonomous driving or medical image diagnosis. However, although the task of confidence calibration has been investigated on classification problems, thorough investigations on object detection and segmentation problems are still missing. Therefore, we focus on the investigation of confidence calibration for object detection and segmentation models in this chapter. We introduce the concept of multivariate confidence calibration that is an extension of well-known calibration methods to the task of object detection and segmentation. This allows for an extended confidence calibration that is also aware of additional features such as bounding box/pixel position, shape information, etc. Furthermore, we extend the expected calibration error (ECE) to measure miscalibration of object detection and segmentation models. We examine several network architectures on MS COCO as well as on Cityscapes and show that especially object detection as well as instance segmentation models are intrinsically miscalibrated given the introduced definition of calibration. Using our proposed calibration methods, we have been able to improve calibration so that it also has a positive impact on the quality of segmentation masks as well.
Most existing real-time deep models trained with each frame independently may produce inconsistent results across the temporal axis when tested on a video sequence. A few methods take the correlations in the video sequence into account,e.g., by propagating the results to the neighboring frames using optical flow or extracting frame representations using multi-frame information, which may lead to inaccurate results or unbalanced latency. In this work, we focus on improving the temporal consistency without introducing computation overhead in inference. To this end, we perform inference at each frame. Temporal consistency is achieved by learning from video frames with extra constraints during the training phase. introduced for inference. We propose several techniques to learn from the video sequence, including a temporal consistency loss and online/offline knowledge distillation methods. On the task of semantic video segmentation, weighing among accuracy, temporal smoothness, and efficiency, our proposed method outperforms keyframe-based methods and a few baseline methods that are trained with each frame independently, on datasets including Cityscapes, Camvid, and 300VW-Mask. We further apply our training method to video instance segmentation on YouTubeVISand develop an application of portrait matting in video sequences, by segmenting temporally consistent instance-level trimaps across frames. Experiments show superior qualitative and quantitative results. Code is available at: https://git.io/vidseg.
Most instance segmentation models are not end-to-end trainable due to either the incorporation of proposal estimation (RPN) as a pre-processing or non-maximum suppression (NMS) as a post-processing. Here we propose a novel end-to-end instance segmentation method termed ISDA. It reshapes the task into predicting a set of object masks, which are generated via traditional convolution operation with learned position-aware kernels and features of objects. Such kernels and features are learned by leveraging a deformable attention network with multi-scale representation. Thanks to the introduced set-prediction mechanism, the proposed method is NMS-free. Empirically, ISDA outperforms Mask R-CNN (the strong baseline) by 2.6 points on MS-COCO, and achieves leading performance compared with recent models. Code will be available soon.