際際滷

際際滷Share a Scribd company logo

LIAO TSEN YUNG Resume

?
?
Tsen Yung Liao
Tsen Yung Liao

This resume summarizes the work experience and qualifications of Tsen-Yun Leo. Some key points include: - Over 10 years of experience developing computer vision and machine learning products including IP camera modules, AOI inspection tools, and an intelligent baby monitor. - Expertise in areas such as CUDA optimization, GPU programming, computer vision algorithms, and machine learning techniques. - Educational background includes a Ph.D. in Communications, MS in Communications and Signal Processing, and BS in Electrical Engineering and Computer Science.

1 of 1
Download to read offline
LIA0,TSEN-YUNG ( LEO ) RESUME VISUALIZATION 2010 IP-CAM PEDESTRIAN TRACKING MODULE 2011 SECUTECH BEST- AWAED IP -CAM PRODUCT 2012 MAKE AOI CUDA INSPECTION & MEASUREMENT 2013 SPEED-UP OPTIMIZATION ON ALL AOI PRODUCTS 2014 2D -> 3D INSPECTION & MACHINE LEARNING PRODUCT LINE MODULARIZED PLUG-INS PIPELINE ACCURACY/PRECISION MEASUREMENT INTER/ INTRA INSPECTION GPGPU LATENCY HIDING TECHNIQUES GPU 慌ah圻t: 1. ?繁今靜g 2. 電屁?R, 閲窒音駅勣議輙 3. 恂並勣?匯崑, 厚嶷勣議恂並際輩 2015 INTELLIGENT BABY MONITOR PRODUCT PhDCandidate CommunicationDepartment Military 2006-2008 MS.communicationandSignalprocessing 2004-2006 EEBS2000-2004 EDUCATION @ NTU EE flightingtrainschool short-termprogram Robust monotonic sequence algorithm on emboss defeat inspection Orientation-range color filtering for inspection Sub-pixel measurement via Facet Model/Gaussian N-jet model CNY OCR module:?Lightweight DSP SVM / LDA QR/DataMatrix Decoder and Qualification Build CUDA?Test-suite Validation ( Precision and Regression Unit Test) CUDA 5 profile-based optimization? Reduce memory traffic and meet peak BW, balance between?TLP and?ILP in CUDA kernel fusion way More CUDA Streams asynchronous operation / Texture object Design?CUDA memory pool ( Object pool ) FFT-based Image RegistrationBilateral NLM De-noising pre-processing CUDA Run-length compression/decompression for saving global memory REGION MAP Transparent Glue-over/less Visual Inspection Integration with open-source ?CUB project Dim defeat inspection on blank Full-Sheet Texture analysis/synthesis?Two-pass module for pattern full-sheet ? Context analysis: Compressed?domain?Video?Saliency?Map QC tool:?Color fiducial mark/ Camera + Line Scan Camera FOV Pose / Belt Line Vibration / Benchmark Test ? Shape-context OCR module : TPS and K-D tree Sub-pixel Measurement Super-Resolution Bar-code Decoder FAST-corner features plus with DTW ( Dynamic Time Warping) ?matching Textile Inspection: CUDA Dynamic parallelism?Kernel ? Stereoscopic Inspection : CUDA KinectFusion ( SLAM ) Dynamic Inspection ROI Generator ARM-based CUDA Toolkit ( Tegra K1 GPGPU ?)? 2013 2014 2012 Recti?cation ? Registration Regularization Restoration Relaxation Insight Learning App UserCloud KNOW-HOW EXPERIENCE AOI Product Dynamic ROI Generator Segmentation = ? Classi?cation?= Reconstruction OCR ML Classifier Intelligent Baby Monitor Gst CryClassi?er LaSVM Gst FaceDetector AdaBoost Gst FacetTypeClassi?er? CNN ?Deep-learning Gst SleepDetector Adaptive Skin Model? Gst InCrib Detector P-expert /N-expert Reinforcement learning on-line incremental learning Cross-validation Over-?tting v.s. Generalization 15定5?埖17?晩佛豚?晩

More Related Content

LIAO TSEN YUNG Resume

  • 1. LIA0,TSEN-YUNG ( LEO ) RESUME VISUALIZATION 2010 IP-CAM PEDESTRIAN TRACKING MODULE 2011 SECUTECH BEST- AWAED IP -CAM PRODUCT 2012 MAKE AOI CUDA INSPECTION & MEASUREMENT 2013 SPEED-UP OPTIMIZATION ON ALL AOI PRODUCTS 2014 2D -> 3D INSPECTION & MACHINE LEARNING PRODUCT LINE MODULARIZED PLUG-INS PIPELINE ACCURACY/PRECISION MEASUREMENT INTER/ INTRA INSPECTION GPGPU LATENCY HIDING TECHNIQUES GPU 慌ah圻t: 1. ?繁今靜g 2. 電屁?R, 閲窒音駅勣議輙 3. 恂並勣?匯崑, 厚嶷勣議恂並際輩 2015 INTELLIGENT BABY MONITOR PRODUCT PhDCandidate CommunicationDepartment Military 2006-2008 MS.communicationandSignalprocessing 2004-2006 EEBS2000-2004 EDUCATION @ NTU EE flightingtrainschool short-termprogram Robust monotonic sequence algorithm on emboss defeat inspection Orientation-range color filtering for inspection Sub-pixel measurement via Facet Model/Gaussian N-jet model CNY OCR module:?Lightweight DSP SVM / LDA QR/DataMatrix Decoder and Qualification Build CUDA?Test-suite Validation ( Precision and Regression Unit Test) CUDA 5 profile-based optimization? Reduce memory traffic and meet peak BW, balance between?TLP and?ILP in CUDA kernel fusion way More CUDA Streams asynchronous operation / Texture object Design?CUDA memory pool ( Object pool ) FFT-based Image RegistrationBilateral NLM De-noising pre-processing CUDA Run-length compression/decompression for saving global memory REGION MAP Transparent Glue-over/less Visual Inspection Integration with open-source ?CUB project Dim defeat inspection on blank Full-Sheet Texture analysis/synthesis?Two-pass module for pattern full-sheet ? Context analysis: Compressed?domain?Video?Saliency?Map QC tool:?Color fiducial mark/ Camera + Line Scan Camera FOV Pose / Belt Line Vibration / Benchmark Test ? Shape-context OCR module : TPS and K-D tree Sub-pixel Measurement Super-Resolution Bar-code Decoder FAST-corner features plus with DTW ( Dynamic Time Warping) ?matching Textile Inspection: CUDA Dynamic parallelism?Kernel ? Stereoscopic Inspection : CUDA KinectFusion ( SLAM ) Dynamic Inspection ROI Generator ARM-based CUDA Toolkit ( Tegra K1 GPGPU ?)? 2013 2014 2012 Recti?cation ? Registration Regularization Restoration Relaxation Insight Learning App UserCloud KNOW-HOW EXPERIENCE AOI Product Dynamic ROI Generator Segmentation = ? Classi?cation?= Reconstruction OCR ML Classifier Intelligent Baby Monitor Gst CryClassi?er LaSVM Gst FaceDetector AdaBoost Gst FacetTypeClassi?er? CNN ?Deep-learning Gst SleepDetector Adaptive Skin Model? Gst InCrib Detector P-expert /N-expert Reinforcement learning on-line incremental learning Cross-validation Over-?tting v.s. Generalization 15定5?埖17?晩佛豚?晩