Internship Decoder Jobs (NOW HIRING)

About Ofinno: 
Ofinno is a leading research and development lab headquartered in Reston, Virginia, specializing in advancing communication and media standards. Our team’s innovative work has led to significant contributions to technologies such as 5G cellular, Wi-Fi, and media compression. Ofinno holds strategic partnerships and licensing agreements with several of the world’s leading technology companies that use such technologies. At Ofinno, we foster an environment of collaboration and excellence, where researchers can focus on delivering breakthroughs that shape the future of technology. 
Position Overview:
As a Summer Research Intern in the Advanced Media Lab, you will work alongside two senior engineers to jump-start Ofinno’s 3DSDK software stack around 3D Gaussian Splatting (3DGS) coding. The 10-week, on-site summer internship is software-integration-first, with research-flavored work layered on top as time allows. You will integrate V-PCC and 3DGS preprocessing into ffmpeg, contribute to an initial encoder-plus-decoder tool chain, and ship reproducible code that the AML team can iterate on after you leave.
Key Responsibilities:
As a Summer Research Intern in 3D Gaussian Splatting, you will:

• Integrate MPEG V-PCC into ffmpeg — encoder-side wrapper / filter / codec module, build-system integration, and test fixtures against reference content.
• Implement a basic 3DGS preprocessing path in ffmpeg — a clean preprocessing utility that takes a reference Gaussian Splat asset and produces the input expected by the encoder.
• Contribute to an initial encoder-plus-decoder tool chain — the decoder-side counterpart and the glue needed to run an end-to-end encode → decode round-trip on at least one reference asset.
• Make integration-level design and trade-off calls — parameter defaults, container choices, color and colorspace handling, basic profiling and optimization.
• Document and hand off the work — deliver a week-3 integration plan, a week-6 mid-program demo, and a final technical report plus handoff doc supporting FTE follow-on at week 10.
• If time allows, pursue a scoped research-flavored exploration informed by integration learnings — e.g., a preprocessing variation, an RD trade-off study, or a profiling-driven optimization.

Qualifications:
Minimum:

• Currently enrolled Ph.D. student (post-qualifying-exam) in EE / CS or related field — graphics, computer vision, signal processing, or systems.
• Strong C / C++ — comfortable reading and modifying a non-trivial production codebase (ffmpeg, V-PCC reference implementation, or equivalent).
• Python proficiency for tooling, scripting, and tests.
• Familiarity with at least one of: video / point-cloud / 3DGS coding stacks; ffmpeg internals (libavcodec, libavformat, the filter API); build systems for large C / C++ projects.
• Engineering judgment — willing to choose pragmatic integration paths over open-ended research detours, and to scope work to fit a 10-week window.
• Strong written and verbal communication, including the ability to document design decisions, write a clear technical report, and present results to the team.

Preferred:

• Prior experience touching a real codec or media pipeline — school project, open-source contribution, prior internship, or thesis work.
• Exposure to 3D Gaussian Splatting, point-cloud compression, neural rendering, or learned image / video compression.
• Comfortable with profiling tools (perf, VTune, gprof, or equivalent) and basic performance work on C / C++ codebases.
• Familiarity with rate-distortion methodology and standard quality metrics.

Powered by JazzHR

qlTGhbyoVL