AI/ML · Computer Vision · Research/ADAS
Bold tech + smooth motion portfolio.
Placeholder copy for now — we’ll tune identity later based on role. The design + motion foundation comes first.
Now
Sim-to-real merger detection (Thesis)
Training with scarce real labels using domain adaptation (DANN) + multimodal fusion.
See thesis work →Next
Interactive demos on-site
Plugging Hugging Face models into the portfolio so visitors can test classifiers & vision pipelines live.
Browse demos →Proof
- • Real-time perception stack (fusion + tracking + forecasting)
- • Edge/GPU optimization mindset (TensorRT / throughput)
- • Hardware depth: RISC-V + FPGA accelerator build
Explore by Domain
Pick a focus area to see the most relevant projects and case studies.
Perception / ADAS
Safety-focused Perception/ADAS: tracking + trajectory + risk forecasting.
View →
GPU / Systems
Optimization, acceleration, edge constraints, and deployment thinking.
View →
Astrophysics ML
Scientific ML: galaxy mergers, domain shift, scarce labels.
View →
Agents / LLM
Agentic systems, tools, orchestration, and interactive demos.
View →
Vision
Multi-camera fusion, detection, tracking, and real-time pipelines.
View →
Perception / ADAS
Safety-focused Perception/ADAS: tracking + trajectory + risk forecasting.
View →
GPU / Systems
Optimization, acceleration, edge constraints, and deployment thinking.
View →
Astrophysics ML
Scientific ML: galaxy mergers, domain shift, scarce labels.
View →
Agents / LLM
Agentic systems, tools, orchestration, and interactive demos.
View →
Vision
Multi-camera fusion, detection, tracking, and real-time pipelines.
View →