The Lagrange Research Center is exclusively dedicated to fundamental research on the foundations of Mathematics, Computing and Data Science as well as to expand the horizons of these fields by exploring other scientific disciplines through a computational and mathematics lens. The main mission of the Lagrange Research Center is to help to push the boundaries of science in this field and prepare the breakthrough innovations of tomorrow. The insights gained from such explorations will open new directions and advancing our understanding of fundamental issues. All generated research results and organized scientific events are open. The center was built on a unique innovative structure model for industry, based on open long-term research grants that support pioneering projects for top scientists.

You will work in support of the Lagrange team and we proposed the research topics as below:

*Each topic has 3 sub-directions to choose

1. Mathematical Problems in Computer Graphics:

a) Speedup for computational geometry operations (find convex hulls, finding nearest neighbors to a given query point, determining the intersection between two surfaces, and triangulating a polygon ...)

b) Latent representation for models

c) Efficient sampling methods for integral equation

2. Searching for High-efficiency Algorithm of Neural Network Computations Adaptive to Davinci Architecture:

a) Computation form (conv, gemm)

b) Precision (integer, mixed precision, float)

c) Sparsity

3. Quantization and Model Compression in Computer Vision (and Multi-task):

a) Computation analysis for large models (nerf, clip, ...)

b) Co-optimization with hardware (quantization, pruning, low-rank approximation, distillation)

c) Insight for hardware design

4. Research in Programming Paradigm:

a) Review of novel and useful characteristics of programming language (go, rust, swift...)

b) Study of modern multi-processor scheduler (GMP model, co-routine, user space mapping, fairness, waiting queue, lifecycle management)

c) Data-oriented and object-oriented programming, compiling and scheduling

Profiles with the following research backgrounds are preferred:

Ph. D. in mathematics / artificial intelligence / computer and other related professional background from domestic or abroad university; Engaged in optimization, supervision, self-supervision, semi-supervised learning or other related research in computer vision, intelligent speech/semantic processing, multimodal information fusion, compiler optimization, computer architecture design and acceleration or other related fields; Have a deep academic accumulation, with good academic literacy and exploration ability; Proficient spoken and written English, French (Chinese will be a plus) Optional:

Experience in on-device training related projects; Outstanding academic achievements during the doctorate period.

To apply this position, please send us a complete CV and briefly explaining your expertise and research interests to wanglei757@huawei.com

HUAWEI Technologies France SASU (“HUAWEI”) respects your privacy and is committed to the protection of your personal data. Your personal data included in the application may be processed by HUAWEI. With your application, you agree with our Recruitment Privacy Statement: http://career.huawei.com/reccampportal/portal/hrd/weu_rec_all.html