25 the Research

The second year of his master's program is devoted to research, and he finally chose breath tracking using an RGB‑D camera. Since then, he has been dealing with many ideas in image processing and human–computer interaction, as well as a lot of dirty coding and frequent experiments tuning parameters, building a hill of shit in Python and C#.

Following are some very short notes on some paper.

• Candidate Topics

  • Cloud-Device Collaborative Learning via small and large lanauge models for mobile health
  • Efficient Cloth Try-on Image Synthesis via Denoising Diffusion Model
  • Brain-inspired computing by assemblies of neurons
  • Breath Tracking using a Depth Camera for XV Scanner
  • Images to Text for Museum Collection
  • Model Reprogramming for Low-Level Image Processing

• Computer Graphics

  • Guided Image Filtering
  • Motion Magnification

• Computer Vision

  • Motion Magnification
    Bilateral Video Magnification FilterEulerian Video Magnification for Revealing Subtle Changes in the WorldLearning-based Axial Video Motion Magnification
  • PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
  • Remote Photo Plethysmography (rPPG)
    Ballistocardiogram signal processing: a reviewEVM-CNN: Real-Time Contactless Heart Rate Estimation from Facial VideoFeasibility of Heart Rate and Respiratory Rate Estimation by Inertial Sensors Embedded in a Virtual Reality HeadsetPhysFormer++: Facial Video-based Physiological Measurement with SlowFast Temporal Difference TransformerRhythmMamba: Fast, Lightweight, and Accurate Remote Physiological MeasurementrPPG-Toolbox: Deep Remote PPG Toolbox
  • Respiratory Monitoring
    An Unobtrusive and Non-Contact Method for Respiratory Measurement With Respiratory Region Detecting Algorithm Based on Depth ImagesDeepBreath: Breathing Exercise Assessment with a Depth CameraMarkerless Active Trunk Shape Modelling for Motion Tolerant Remote Respiratory AssessmentRemote Pulmonary Function Testing using a Depth SensorSensing Systems for Respiration Monitoring: A Technical Systematic ReviewSoleimani's DatasetVideo-Based Respiratory Rate Estimation for Infants in the NICU
  • Semantic Segmentation
    BiSeNet: Bilateral Segmentation Network for Real-time Semantic SegmentationICNet for Real-Time Semantic Segmentation on High-Resolution ImagesIndoor Semantic Segmentation using depth informationSAMURAI: Adapting Segment Anything Model for Zero-Shot Visual Tracking with Motion-Aware MemorySegment Anything

• Human-Computer Interaction