A5091700001- Advanced Vision and Imaging
- Advanced Image Processing Techniques
- Image Processing Techniques and Applications
- Smart Agriculture and AI
- Image Enhancement Techniques
- Prostate Cancer Diagnosis and Treatment
- Radiomics and Machine Learning in Medical Imaging
- Remote Sensing in Agriculture
- Video Surveillance and Tracking Methods
- Advanced Image Fusion Techniques
- Spectroscopy and Chemometric Analyses
- Date Palm Research Studies
- Visual Attention and Saliency Detection
- Medical Imaging and Analysis
- Remote-Sensing Image Classification
- Advanced X-ray and CT Imaging
- Plant Virus Research Studies
- Advanced Optical Sensing Technologies
- Image and Signal Denoising Methods
- MRI in cancer diagnosis
- Advanced Image and Video Retrieval Techniques
- Advanced Neural Network Applications
- Leaf Properties and Growth Measurement
- Plant Disease Management Techniques
- Optical measurement and interference techniques
Indian Institute of Technology Ropar
2019-2024
Sanskriti Samvardhan Mandal
2019-2020
Due to the unavailability of large scale underwater depth image datasets and ill-posed problems, single prediction is a challenging task. An unambiguous for an essential part applications like robotics, marine engineering, etc. This paper presents end-to-end Underwater Generative Adversarial Network (UW-GAN) estimation from image. Initially, coarse-level map estimated using Coarse-level (UWC-Net). Then, fine-level computed Fine-level (UWF-Net) which takes input as concatenation The proposed...
Plant disease detection and early treatment are essential for sustainable crop production. Computer vision science is overgrowing with the advancement in deep learning. Real time plant poses a challenge due to unpredictable spread of diseases within plant, environmental factors, scarcity real field datasets. The proposed work systematically addresses these issues through three key components: (a) Collaboratively generating novel pigeon pea image dataset from agricultural fields, partnership...
Underwater image restoration is a challenging problem due to the multiple distortions. Degradation in information mainly 1) light scattering effect 2) wavelength dependent color attenuation and 3) object blurriness effect. In this letter, we propose novel end-to-end deep network for underwater restoration. The proposed divided into two parts viz. channel-wise feature extraction module dense-residual module. A custom loss function proposed, which preserves structural details generates true...
Depth prediction from single image is a challenging task due to the intra scale ambiguity and unavailability of prior information. The an unambiguous depth RGB very important aspect for computer vision applications. In this paper, end-to-end sparse-to-dense network (S2DNet) proposed estimation (SIDE). processes along with additional sparse samples estimation. sample are acquired either low-resolution sensor or calculated by visual simultaneous localization mapping (SLAM) algorithms. first...
The quality of images captured in bad weather is often affected by chromatic casts and low visibility due to the presence atmospheric particles. Restoration color balance ignored most existing image de-hazing methods. In this paper, we propose a varicolored end-to-end network which restores given hazy recovers haze-free image. proposed comprises 1) Haze correction (HCC) module 2) Visibility improvement (VI) module. HCC provides required attention each channel generates balanced While VI...
The multi-modality sensor fusion technique is an active research area in scene understating. In this work, we explore the RGB image and semantic-map methods for depth estimation. LiDARs, Kinect, TOF sensors are unable to predict depth-map at illuminate monotonous pattern surface. paper, propose a semantic-to-depth generative adversarial network (S2D-GAN) estimation from its semantic-map. first stage, proposed S2D-GAN estimates coarse level depthmap using semantic-to-coarse-depth (S2CD-GAN)...
Recently, the potential for wheat head detection has been significantly enhanced using deep learning techniques. However, significant challenges are variation in growth stages of heads, canopy, genotype, and orientation. Furthermore, task gets even more complex due to overlapping density heads blur image wind. For real-time detection, designing lightweight models edge devices is also challenging. This paper proposes a WheatNet-Lite architecture enhance efficiency accuracy detection. The...
Scene understanding is an active area of research in computer vision that encompasses several different problems. The LiDARs and stereo depth sensor have their own restrictions such as light sensitiveness, power consumption short-range [1]. In this paper, we propose a two-stream deep adversarial network for single image estimation RGB images. For stream I network, novel encoder-decoder architecture using residual concepts to extract course-level features. Stream II purely processes the...
Plant disease detection and early treatment are essential for sustainable crop production. Computer vision science is overgrowing with the advancement in deep learning. However, plant challenging due to random spread throughout body, environmental challenges, lack of actual field datasets. The proposed work systematically addresses these issues with(a) construction first pigeon pea image dataset from agriculture collaboration 20 Agricultural Research Centers (ARS) government agencies across...
Due to the numerous applications of boundary maps and occlusion orientation (ORI-maps) in high-level vision problems, accurate estimation these is a crucial task. The existing deep networks employ single-stream network estimate relation between map ORI-map estimation. However, fail explore significant individual information separately. To resolve this problem, paper, we propose novel two-stream generative adversarial (GAN) for estimation, named OBP-GAN. proposed OBP-GAN consists two streams...
The increasing demand for computational photography and imaging on mobile platforms has led to the widespread development integration of advanced image sensors with novel algorithms in camera systems. However, scarcity high-quality data research rare opportunity in-depth exchange views from industry academia constrain intelligent (MIPI). Building achievements previous MIPI Workshops held at ECCV 2022 CVPR 2023, we introduce our third challenge including three tracks focusing algorithms. In...
Automated video-based applications are a highly demanding technique from security perspective, where detection of moving objects i.e., object segmentation (MOS) is performed. Therefore, we have proposed an effective solution with spatio-temporal squeeze excitation mechanism (SqEm) based multi-level feature sharing encoder-decoder network for MOS. Here, the SqEm module to get prominent foreground edge information using features. Further, residual decoder respective features and previous...