Search

A wide variety of computer vision applications rely on superpixel or supervoxel algorithms as a preprocessing step. This underlines the overall importance that these approaches have gained in recent years. However, most methods show a lack of temporal consistency or fail in producing temporally stable superpixels. In this paper, we present an approach to generate temporally consistent superpixels for video content. Our method is formulated as a contour-evolving expectation-maximization framework, which utilizes an efficient label propagation scheme to encourage the preservation of superpixel shapes and their relative positioning over time. By explicitly detecting the occlusion of superpixels and the disocclusion of new image regions, our framework is able to terminate and create superpixels whose corresponding image region becomes hidden or newly appears. Additionally, the occluded parts of superpixels are incorporated in the further optimization. This increases the compliance of the superpixel flow with the optical flow present in the scene. Using established benchmark suites, we show the performance of our approach in comparison to state-of-the-art supervoxel and superpixel algorithms for video content.

From January 1980 to December 1989, 86 patients with limited small cell lung carcinoma (SCLC) were treated in our Institution. Sixty-eight of them were males, 18 females; mean age was 58 years (range: 40-74); Karnofsky index was 80 or higher. All patients received induction multiagent chemotherapy (CAV or CAVE), usually fractionated into 3/4 cycles. Radiotherapy was administered by means of a 10 MV Linear Accelerator, 2-3 Gy/day, 5 days/week. Radiation dose was 60 Gy for 39 patients, 50 Gy for 25 patients, and 30 Gy for 22 patients. Parallel opposed fields (AP and PA) were used for administering the 30-Gy dose, while higher doses were delivered by multiportal arrangement (3/4 fields). Overall survival (corrected Kaplan-Meier) was 21.3% and 13.4% at 2 and 3 years, respectively; 2 years' survival according to dose was 24.5% for 60 Gy, 19.9% and 11% for 50 and 30 Gy, respectively. Mean survival time (MST) was 14 months for 60 Gy, 13 months and 10 months, respectively, for 50 and 30 Gy. Survival and disease-free survival rates were similar in patients who received 50 and 60 Gy. No major complications related to irradiation were observed, not even in the group receiving the highest dose. Our data are similar to those reported in the literature.