MIPI 2023 Challenge on RGB+ToF Depth Completion: Methods and Results

no code implementations • 27 Apr 2023 • Qingpeng Zhu, Wenxiu Sun, Yuekun Dai, Chongyi Li, Shangchen Zhou, Ruicheng Feng, Qianhui Sun, Chen Change Loy, Jinwei Gu, Yi Yu, Yangke Huang, Kang Zhang, Meiya Chen, Yu Wang, Yongchao Li, Hao Jiang, Amrit Kumar Muduli, Vikash Kumar, Kunal Swami, Pankaj Kumar Bajpai, Yunchao Ma, Jiajun Xiao, Zhi Ling

To evaluate the performance of different depth completion methods, we organized an RGB+sparse ToF depth completion competition.

Depth Completion

AcED: Accurate and Edge-consistent Monocular Depth Estimation

no code implementations • 16 Jun 2020 • Kunal Swami, Prasanna Vishnu Bondada, Pankaj Kumar Bajpai

Single image depth estimation is a challenging problem.

Monocular Depth Estimation regression

DISCO: Depth Inference from Stereo using Context

no code implementations • 31 May 2019 • Kunal Swami, Kaushik Raghavan, Nikhilanj Pelluri, Rituparna Sarkar, Pankaj Bajpai

Recent deep learning based approaches have outperformed classical stereo matching methods.

Decoder Stereo Matching +1

CANDY: Conditional Adversarial Networks based Fully End-to-End System for Single Image Haze Removal

no code implementations • 9 Jan 2018 • Kunal Swami, Saikat Kumar Das

In this paper, we present CANDY (Conditional Adversarial Networks based Dehazing of hazY images), a fully end-to-end model which directly generates a clean haze-free image from a hazy input image.

Single Image Haze Removal

Why my photos look sideways or upside down? Detecting Canonical Orientation of Images using Convolutional Neural Networks

no code implementations • 4 Dec 2017 • Kunal Swami, Pranav P. Deshpande, Gaurav Khandelwal, Ajay Vijayvargiya

Humans use object recognition and contextual scene information to correctly orient images.

Object Recognition Scene Understanding