The field of computer vision is moving towards more robust and accurate methods for novel view synthesis and depth estimation. Researchers are exploring new approaches to address challenges such as occlusions, pose drift, and information scarcity. One notable direction is the use of diffusion models to improve the quality of synthesized views and estimated depths. Another area of focus is the development of more efficient and effective methods for 3D Gaussian Splatting, including the use of incremental joint optimization and robust pose estimation modules. Noteworthy papers in this area include: DMS, which proposes a model-agnostic approach to synthesize novel views along the epipolar direction, guided by directional prompts. LongSplat, which introduces a robust unposed 3D Gaussian Splatting framework featuring incremental joint optimization and a robust pose estimation module. GSFix3D, which improves the visual fidelity in under-constrained regions by distilling prior knowledge from diffusion models into 3D representations. Enhancing Novel View Synthesis from extremely sparse views with SfM-free 3D Gaussian Splatting Framework, which proposes a novel SfM-free 3DGS-based method that jointly estimates camera poses and reconstructs 3D scenes from extremely sparse-view inputs.