Deep monocular 3D reconstruction for assisted navigation in bronchoscopy.

PURPOSE: In bronchoschopy, computer vision systems for navigation assistance are an attractive low-cost solution to guide the endoscopist to target peripheral lesions for biopsy and histological analysis. We propose a decoupled deep learning architecture that projects input frames onto the domain of CT renderings, thus allowing offline training from patient-specific CT data. METHODS: A fully convolutional network architecture is implemented on GPU and tested on a phantom dataset involving 32 video sequences and [Formula: see text]60k frames with aligned ground truth and renderings, which is made available as the first public dataset for bronchoscopy navigation. RESULTS: An average estimated depth accuracy of 1.5 mm was obtained, outperforming conventional direct depth estimation from input frames by 60%, and with a computational time of [Formula: see text]30 ms on modern GPUs. Qualitatively, the estimated depth and renderings closely resemble the ground truth. CONCLUSIONS: The proposed method shows a novel architecture to perform real-time monocular depth estimation without losing patient specificity in bronchoscopy. Future work will include integration within SLAM systems and collection of in vivo datasets.

Sub-millisecond timing-jitter-free tuning of parametric dispersion compensator.

We demonstrate sub-millisecond tuning of a prototype parametric tunable dispersion compensator (P-TDC) based on cascaded polarization-diverse four-wave mixing (FWM) process with a fast tunable and highly wavelength-stable pump light source. The pump light source is developed using a tunable distributed amplification chirped sampled grating distributed reflector laser that is fully wavelength tunable by on-chip heaters with a 3-dB frequency response of 45 kHz, resulting in fast dispersion tuning of less than 50 µs without additional timing jitter. The P-TDC is developed as the first prototype to satisfy essential requirements for practical network uses: stable input-polarization diversity, input-wavelength preservation, and seamless dispersion tunability for entire C-band input wavelengths are simultaneously achieved.

Discriminative feature co-occurrence selection for object detection.

This paper describes an object detection framework that learns the discriminative co-occurrence of multiple features. Feature co-occurrences are automatically found by Sequential Forward Selection at each stage of the boosting process. The selected feature co-occurrences are capable of extracting structural similarities of target objects leading to better performance. The proposed method is a generalization of the framework proposed by Viola and Jones, where each weak classifier depends only on a single feature. Experimental results obtained using four object detectors, for finding faces and three different hand gestures, respectively, show that detectors trained with the proposed algorithm yield consistently higher detection rates than those based on their framework while using the same number of features.