The problem of sketch semantic segmentation is far from being solved. Despite existing methods exhibiting near-saturating performances on simple sketches with high recognisability, they suffer ...serious setbacks when the target sketches are products of an imaginative process with high degree of creativity. We hypothesise that human creativity, being highly individualistic, induces a significant shift in distribution of sketches, leading to poor model generalisation. Such hypothesis, backed by empirical evidences, opens the door for a solution that explicitly disentangles creativity while learning sketch representations. We materialise this by crafting a learnable creativity estimator that assigns a scalar score of creativity to each sketch. It follows that we introduce CreativeSeg, a learning-to-learn framework that leverages the estimator in order to learn creativity-agnostic representation, and eventually the downstream semantic segmentation task. We empirically verify the superiority of CreativeSeg on the recent "Creative Birds" and "Creative Creatures" creative sketch datasets. Through a human study, we further strengthen the case that the learned creativity score does indeed have a positive correlation with the subjective creativity of human. Codes are available at https://github.com/PRIS-CV/Sketch-CS .
Human free-hand sketches provide the useful data for studying human perceptual grouping, where the grouping principles such as the Gestalt laws of grouping are naturally in play during both the ...perception and sketching stages. In this paper, we make the first attempt to develop a universal sketch perceptual grouper. That is, a grouper that can be applied to sketches of any category created with any drawing style and ability, to group constituent strokes/segments into semantically meaningful object parts. The first obstacle to achieving this goal is the lack of large-scale datasets with grouping annotation. To overcome this, we contribute the largest sketch perceptual grouping dataset to date, consisting of 20 000 unique sketches evenly distributed over 25 object categories. Furthermore, we propose a novel deep perceptual grouping model learned with both generative and discriminative losses. The generative loss improves the generalization ability of the model, while the discriminative loss guarantees both local and global grouping consistency. Extensive experiments demonstrate that the proposed grouper significantly outperforms the state-of-the-art competitors. In addition, we show that our grouper is useful for a number of sketch analysis tasks, including sketch semantic segmentation, synthesis, and fine-grained sketch-based image retrieval.
Abstract
The purpose of this study was to examine whether the increased risk of colorectal cancer due to cigarette smoking differed by anatomical subsite or sex. We analyzed data from 188,052 ...participants aged 45–75 years (45% men) who were enrolled in the Multiethnic Cohort Study in 1993–1996. During a mean follow-up period of 16.7 years, we identified 4,879 incident cases of invasive colorectal adenocarcinoma. In multivariate Cox regression models, as compared with never smokers of the same sex, male ever smokers had a 39% higher risk (hazard ratio (HR) = 1.39, 95% confidence interval (CI): 1.16, 1.67) of cancer of the left (distal or descending) colon but not of the right (proximal or ascending) colon (HR = 1.03, 95% CI: 0.89, 1.18), while female ever smokers had a 20% higher risk (HR = 1.20, 95% CI: 1.06, 1.36) of cancer of the right colon but not of the left colon (HR = 0.96, 95% CI: 0.80, 1.15). Compared with male smokers, female smokers had a greater increase in risk of rectal cancer with number of pack-years of smoking (P for heterogeneity = 0.03). Our results suggest that male smokers are at increased risk of left colon cancer and female smokers are at increased risk of right colon cancer. Our study also suggests that females who smoke may have a higher risk of rectal cancer due to smoking than their male counterparts.
•Star graph ensemble matching is introduced to address the structure of the sketch.•MKL is employed to reach the state-of-the-art sketch recognition performance.•Sketch attributes are explored for ...sketch recognition and interesting applications.
Free-hand sketch recognition has become increasingly popular due to the recent expansion of portable touchscreen devices. However, the problem is non-trivial due to the complexity of internal structures that leads to intra-class variations, coupled with the sparsity in visual cues that results in inter-class ambiguities. In order to address the structural complexity, a novel structured representation for sketches is proposed to capture the holistic structure of a sketch. Moreover, to overcome the visual cue sparsity problem and therefore achieve state-of-the-art recognition performance, we propose a Multiple Kernel Learning (MKL) framework for sketch recognition, fusing several features common to sketches. We evaluate the performance of all the proposed techniques on the most diverse sketch dataset to date (Mathias et al., 2012), and offer detailed and systematic analyses of the performance of different features and representations, including a breakdown by sketch-super-category. Finally, we investigate the use of attributes as a high-level feature for sketches and show how this complements low-level features for improving recognition performance under the MKL framework, and consequently explore novel applications such as attribute-based retrieval.
In this paper, for the first time, we investigate the problem of generating 3D shapes from professional 2D sketches via deep learning. We target sketches done by professional artists, as these ...sketches are likely to contain more details than the ones produced by novices, and thus the reconstruction from such sketches poses a higher demand on the level of detail in the reconstructed models. This is importantly different to previous work, where the training and testing was conducted on either synthetic sketches or sketches done by novices. Novices sketches often depict shapes that are physically unrealistic, while models trained with synthetic sketches could not cope with the level of abstraction and style found in real sketches. To address this problem, we collected the first large-scale dataset of professional sketches, where each sketch is paired with a reference 3D shape, with a total of 1,500 professional sketches collected across 500 3D shapes. The dataset is available at http://sketchx.ai/downloads/ . We introduce two bespoke designs within a deep adversarial network to tackle the imprecision of human sketches and the unique figure/ground ambiguity problem inherent to sketch-based reconstruction. We show that existing 3D shapes generation methods designed for images fail to be naively applied to our problem, and demonstrate the effectiveness of our method both qualitatively and quantitatively.
Generalized Few-shot Semantic Segmentation (GFSS) aims to segment each image pixel into either base classes with abundant training examples or novel classes with only a handful of ( e . g ., 1-5) ...training images per class. Compared to the widely studied Few-shot Semantic Segmentation (FSS), which is limited to segmenting novel classes only, GFSS is much under-studied despite being more practical. Existing approach to GFSS is based on classifier parameter fusion whereby a newly trained novel class classifier and a pre-trained base class classifier are combined to form a new classifier. As the training data is dominated by base classes, this approach is inevitably biased towards the base classes. In this work, we propose a novel Prediction Calibration Network (PCN) to address this problem. Instead of fusing the classifier parameters, we fuse the scores produced separately by the base and novel classifiers. To ensure that the fused scores are not biased to either the base or novel classes, a new Transformer-based calibration module is introduced. It is known that the lower-level features are useful of detecting edge information in an input image than higher-level features. Thus, we build a cross-attention module that guides the classifier's final prediction using the fused multi-level features. However, transformers are computationally demanding. Crucially, to make the proposed cross-attention module training tractable at the pixel level, this module is designed based on feature-score cross-covariance and episodically trained to be generalizable at inference time. Extensive experiments on PASCAL-<inline-formula> <tex-math notation="LaTeX">5^{i} </tex-math></inline-formula> and COCO-<inline-formula> <tex-math notation="LaTeX">20^{i} </tex-math></inline-formula> show that our PCN outperforms the state-the-the-art alternatives by large margins.
We present the first one-shot personalized sketch segmentation method. We aim to segment all sketches belonging to the same category provisioned with a single sketch with a given part annotation ...while (i) preserving the parts semantics embedded in the exemplar, and (ii) being robust to input style and abstraction. We refer to this scenario as personalized . With that, we importantly enable a much-desired personalization capability for downstream fine-grained sketch analysis tasks. To train a robust segmentation module, we deform the exemplar sketch to each of the available sketches of the same category. Our method generalizes to sketches not observed during training. Our central contribution is a sketch-specific hierarchical deformation network. Given a multi-level sketch-strokes encoding obtained via a graph convolutional network, our method estimates rigid-body transformation from the target to the exemplar, on the upper level. Finer deformation from the exemplar to the globally warped target sketch is further obtained through stroke-wise deformations, on the lower-level. Both levels of deformation are guided by mean squared distances between the keypoints learned without supervision, ensuring that the stroke semantics are preserved. We evaluate our method against the state-of-the-art segmentation and perceptual grouping baselines re-purposed for the one-shot setting and against two few-shot 3D shape segmentation methods. We show that our method outperforms all the alternatives by more than 10% on average. Ablation studies further demonstrate that our method is robust to personalization : changes in input part semantics and style differences.
A common strategy adopted by existing state-of-the-art unsupervised domain adaptation (UDA) methods is to employ two classifiers to identify the misaligned local regions between source and target ...domain. Following the 'wisdom of the crowd' principle, one has to ask: why stop at two? Indeed, we find that using more classifiers leads to better performance, but also introduces more model parameters, therefore risking overfitting. In this paper, we introduce a novel method called STochastic clAssifieRs (STAR) for addressing this problem. Instead of representing one classifier as a weight vector, STAR models it as a Gaussian distribution with its variance representing the inter-classifier discrepancy. With STAR, we can now sample an arbitrary number of classifiers from the distribution, whilst keeping the model size the same as having two classifiers. Extensive experiments demonstrate that a variety of existing UDA methods can greatly benefit from STAR and achieve the state-of-the-art performance on both image classification and semantic segmentation tasks.
The dual transition metals CoS2–SnO2@ reduced graphene oxide heterostructure quantum dots (shorten as CoS2–SnO2@rGO QDs) is intentionally synthesized by hydrothermal method. The compositions, grain ...sizes and contents of the synthesized CoS2–SnO2@rGO QDs are identified and calculated from XRD pattern. The layered structures and crystal sizes (<10 nm) of CoS2–SnO2 heterostructure quantum dots are confirmed by TEM high-resolution image. Moreover, the CoS2–SnO2@rGO QDs deliver higher specific capacity of 940.4, 781.9, 673.8, 506.6 and 303.1 mAh g−1 at 0.2, 0.5, 1.0, 2.0 and 4.0 A g−1, respectively, and remained 521.4 mAh g−1 (300 cycles) with 98.9% Coulomb efficiency under 1.0 A g−1 cycling. The DFT (density functional theory) calculation results suggest that the Li ions should diffuse through two possible paths with relatively lower Ebar values of 0.519 and 0.566 eV in CoS2–SnO2@rGO QDs. Combining with XRD, XPS, TEM and DFT calculation, we therefore assign the excellent electrochemical performances of CoS2–SnO2@rGO QDs to its highly active CoS2–SnO2 interfaces for electrode reactivity enhancement, to the characteristic heterointerfaces for “extra” Li+ ions storage/release, and to super conductive/layered rGO network for better ionic/electronic transport. The successful synthesis of CoS2–SnO2@rGO QDs may provide a strategy for dual/multi-transition metal heterojunction to enhance lithium-ion battery electrode materials.
Display omitted
•Grains of SnO2 and CoS2 in SnO2–CoS2@rGO QDs are ∼3.5 and ∼11.9 nm, respectively.•SnO2–CoS2@rGO QDs remains ∼521.4 mAh/g under 1.0 A/g after 300 cycles.•Ebar of SnO2–CoS2 heterointerface for Li are 0.519 and 0.566 eV by DFT calculation.
Indoor organic photovoltaics (OPVs) are a potential niche application for organic semiconductors due to their strong and well‐matched absorption with the emission of indoor lighting. However, due to ...extremely low photocurrent generation, the device parameters critical for efficient indoor OPVs differ from those under 1 Sun conditions. Herein, these critical device parameters—recombination loss and shunt resistance (Rsh)—are identified and it is demonstrated that bilayer OPVs are suitable for indoor PV applications. Compared to bulk‐heterojunction (BHJ), the open‐circuit voltage loss of bilayer devices under low light intensities is much smaller, consistent with a larger surface photovoltage response, indicating suppressed recombination losses. The bilayer devices show a higher fill factor at low light intensities, resulting from high Rsh afforded by the ideal interfacial contacts between the photoactive and the charge transport layers. The high Rsh enables bilayer devices to perform well without a light‐soaking process. Finally, the charge carriers are extracted rapidly in bilayers, which are attributed to strongly suppressed trap states possibly induced by isolated domains and non‐ideal interfacial contacts in BHJs. This study highlights the excellent suitability of bilayer OPVs for indoor applications and demonstrates the importance of device architecture and interfacial structures for efficient indoor OPVs.
The excellent suitability of bilayer organic photovoltaic devices for efficient indoor‐light harvesting is demonstrated. Thanks to ideal interfacial contacts between photoactive and charge transport layers and minimized isolated donor and/or acceptor domains, bilayer devices show much smaller open‐circuit voltage loss and high shunt resistance, leading to enhanced fill factor and max power output without light soaking.