Existing volumetric neural rendering techniques, such as Neural Radiance Fields (NeRF), face limitations in synthesizing high-quality novel views when the camera poses of input images are imperfect. To address this issue, we propose a novel 3D reconstruction framework that enables simultaneous optimization of camera poses, dubbed CBARF (Cascaded Bundle-Adjusting NeRF). In a nutshell, our framework optimizes camera poses in a coarse-to-fine manner and then reconstructs scenes based on the rectified poses. It is observed that the initialization of camera poses has a significant impact on the performance of bundle-adjustment (BA). Therefore, we cascade multiple BA modules at different scales to progressively improve the camera poses. Meanwhile, we develop a neighbor-replacement strategy to further optimize the results of BA in each stage. In this step, we introduce a novel criterion to effectively identify poorly estimated camera poses. Then we replace them with the poses of neighboring cameras, thus further eliminating the impact of inaccurate camera poses. Once camera poses have been optimized, we employ a density voxel grid to generate high-quality 3D reconstructed scenes and images in novel views. Experimental results demonstrate that our CBARF model achieves state-of-the-art performance in both pose optimization and novel view synthesis, especially in the existence of large camera pose noise.

Inspired by the decision learning process of multilayer per-ceptron (MLP) and decision tree (DT), a new classification model, named the subspace learning machine (SLM), is proposed in this work. SLM first identifies a discriminant subspace, S0, by examining the discriminant power of each input feature. Then, it learns projections of features in S0to yield 1D subspaces and finds the optimal partition for each. A criterion is developed to choose the best q partitions that yield 2qpartitioned subspaces. The partitioning process is recursively applied at each child node to build an SLM tree. When the samples at a child node are sufficiently pure, the partitioning process stops, and each leaf node makes a prediction. The ensembles of SLM trees can yield a stronger predictor. Extensive experiments are conducted for performance benchmarking among SLM trees, ensembles and classical classifiers.

Subspace partitioning in a high-dimensional feature space plays a fundamental role in the design of effective classifiers. A novel subspace learning machine (SLM) that projects high-dimensional feature vectors into a 1D feature subspace and partitions it into two disjoint sets was recently proposed. As an extension, SLM with soft partitioning, denoted by SLM/SP, is proposed in this work. SLM/SP adopts the soft decision tree (SDT) data structure for decision learning. It starts by learning an adaptive tree structure by using local greedy subspace partitioning. Once the stopping criteria are met for all child nodes and the tree structure is determined, all projection vectors are updated globally. This methodology enables efficient training, high classification accuracy, and a small model size. It is shown by experimental results that an SLM/SP tree offers a lightweight and high performance classification method.

Supervised and unsupervised deep trackers that rely on deep learning technologies are popular in recent years. Yet, they demand high computational complexity and a high memory cost. A green unsupervised single-object tracker, called GUSOT, that aims at object tracking for long videos under a resource-constrained environment is proposed in this work. Built upon a baseline tracker, UHP-SOT++, which works well for short-term tracking, GUSOT contains two additional new modules: 1) lost object recovery, and 2) color-saliency-based shape proposal. They help resolve the tracking loss problem and offer a more flexible object proposal, respectively. Thus, they enable GUSOT to achieve higher tracking accuracy in the long run. We conduct experiments on the large-scale dataset LaSOT with long video sequences, and show that GUSOT offers a lightweight high-performance tracking solution that finds applications in mobile and edge computing platforms.

An unsupervised online object tracking method that exploits both foreground and background correlations is proposed and named UHP-SOT (Unsupervised High-Performance Single Object Tracker) in this work. UHP-SOT consists of three modules: 1) appearance model update, 2) background motion modeling, and 3) trajectory-based box prediction. A state-of-the-art discriminative correlation filters (DCF) based tracker is adopted by UHP-SOT as the first module. We point out shortcomings of using the first module alone such as failure in recovering from tracking loss and inflexibility in object box adaptation and then propose the second and third modules to overcome them. Both are novel in single object tracking (SOT). We test UHP-SOT on two popular object tracking benchmarks, TB-50 and TB-100, and show that it outperforms all previous unsupervised SOT methods, achieves a performance comparable with the best supervised deep-learning-based SOT methods, and operates at a fast speed (i.e. 22.7-32.0 FPS on a CPU).