Search

Image scaling techniques such as Super-Resolution (SR) are useful for object detection, especially for detecting small objects. However, we found that scaling by an inappropriate factor tends to induce false-positive detections. This paper presents a Region-Dependent Scale-Proposal (RDSP) network that estimates the appropriate scale factors for each image region depending on its contextual information. In our detection framework, images are appropriately scaled by SR according to the estimations of the RDSP network, and fed into the scale-specific object detectors. While previous works have proposed models for scale proposal, our RDSP extracts regions where objects could potentially exist based on scene structure, regardless of whether actual objects are present, because small objects are often too small to determine their presence accurately. Additionally, while existing approaches have fused object detection and SR in an end-to-end manner, scale proposals for SR are not provided or are performed independently. Qualitative and quantitative experiments show that our RDSP network provides appropriate SR scales and improve detection accuracy on highly challenging dataset, captured by real car-mounted cameras with size-varied objects, including extremely small objects. Our code is available at https://github.com/kakitamedia/RDSP.

Abstract Background DNA methylation in cancer is considered a diagnostic and predictive biomarker. We investigated the usefulness of the methylation status of CALN1 as a biomarker for bladder cancer using methylation-sensitive restriction enzyme (MSRE)-quantitative polymerase chain reaction (qPCR). Methods Eighty-two bladder cancer fresh samples were collected via transurethral resection of bladder tumors. Genomic DNA was extracted from the samples, and MSRE-qPCR was performed to determine the CALN1 methylation percentage. Reverse transcription-qPCR was performed to assess the correlation between CALN1 methylation and mRNA expression. The association between CALN1 methylation percentage and clinicopathological variables of all cases and intravesical recurrence of non-muscle-invasive bladder cancer (non-MIBC) cases were analyzed. Results Of the 82 patients, nine had MIBC and 71 had non-MIBC who had not undergone total cystectomy. The median CALN1 methylation percentage was 79.5% (interquartile range: 51.1–92.6%). The CALN1 methylation percentage had a negative relationship with CALN1 mRNA expression (Spearman’s ρ = − 0.563 and P = 0.012). Hypomethylation of CALN1 was associated with advanced tumor stage (P = 0.0007) and histologically high grade (P = 0.018). Furthermore, multivariate analysis revealed that CALN1 hypomethylation was an independent risk factor for intravesical recurrence in non-MIBC patients (hazard ratio 3.83, 95% confidence interval; 1.14–13.0, P = 0.031). Conclusion Our findings suggest that CALN1 methylation percentage could be a useful molecular biomarker for bladder cancer.