Shuran Song will present her Generals on 5/1/15 at 10:30am in CS 301. Her committee members are Jianxiong Xiao (adviser), Thomas Funkhouser, and Adam Finkelstein. Everyone is invited to attend her talk, and those faculty wishing to remain for the oral exam following are welcome to do so. Her abstract and reading list follow below. Abstract: Object recognition in visual scenes is one of the most fundamental tasks in computer vision. Despite significant progress, object recognition is still a challenging task due to the variations of texture, illumination, shape, viewpoint, clutter, and occlusion. Recently, inexpensive depth cameras have become widely available, and the depth information from these sensors has greatly simplified some common challenges in computer vision. Also, various crowd-sourcing efforts in recent years have resulted in several large-scale on-line 3D CAD repositories. The combination of low-cost RGB-D sensors and large on-line 3D model repositories has a huge potential to revolutionize object recognition in 3D. My thesis will focus on exploiting depth information to address main difficulties in object detection using a novel 3D reasoning framework, together with data-driven 3D features using a huge amount of RGB-D and 3D data. Textbook : Computer Vision: Algorithms and Applications Paper: [DalalTriggs] Histograms of oriented gradients for human detection. [DPM] Object Detection with Discriminatively Trained Part Based Models. [ExemplarSVMs] Ensemble of exemplar-svms for object detection and beyond. [SIFT] Distinctive image features from scale-invariant keypoints. [SelectiveSearch] Segmentation as selective search for object recognition. [Kinect human Pose] Real-time human pose recognition in parts from single depth images. Jamie Shotton, Toby Sharp, Alex Kipman, Andrew Fitzgibbon, Mark Finocchio, Andrew Blake, Mat Cook, andRichard Moore. Communications of the ACM, 2013. [KinectFusion] KinectFusion: Real-time dense surface mapping and tracking. [DepthRCNN] Learning Rich Features from RGB-D Images for Object Detection and Segmentation. Saurabh Gupta, Ross Girshick, Pablo Arbeláez, Jitendra Malik [SUN3D] SUN3D: A Database of Big Spaces Reconstructed using SfM and Object Labels. [PASCAL] The pascal visual object classes (voc) challenge. [ImageNet] Imagenet: A large-scale hierarchical image database.