cs7380 Special Topics in Graphics/Image Processing has a topic on applied geometric representation and computation. The course schedule has quite some pointers to up-to-date reading materials. It is given by Marsette Vona.
Try to compile PTAM on 32-bit Fedora 11. The process goes quite smooth. The only manual effort is to include a missing header file into the source code and to modify a few parameters in Makefile.
OpenKinect, open source project for Kinect.
Proceedings of IEEE Dec 2010 has a special issue on Nanoelectronics Research for Beyond CMOS Information Processing.
Some interesting preprints in TPDS:
- Automatic Generation of Multicore Chemical Kernels
- Comparing Hardware Accelerators in Scientific Applications: A Case Study
- A Framework for Evaluating High-Level Design Methodologies for High-Performance Reconfigurable Computers
- Accelerating the Execution of Matrix Languages on the Cell Broadband Engine Architecture
- Design and Performance Evaluation of Image Processing Algorithms on GPUs
- Exploiting Memory Access Patterns to Improve Memory Performance in Data-Parallel Architectures
Some interesting papers in Annuals of the History of Computing Oct.-Dec. 2010:
- The Inconsistent Youth of Charles Babbage
- From Art Form to Engineering Discipline? A History of US Military Software Development Standards, 1974-1998. Lessons learned form history.
- The MIL MF7114 Microprocessor, the amazing starting point of a golden age.
- John R. Rice: Mathematical Software Pioneer
- RAID: A Personal Recollection of How Storage Became a System
Some interesting papers in TCAD Dec. 2010:
- An Analytical Approach for Network-on-Chip Performance Analysis
- Buffer Optimization in Network-on-Chip Through Flow Regulation
Reconfigurable system for real-time embedded control applications, in IET-CTA 2010.
Some interesting papers in Computer Architecture Letters:
- ParMiBench – An Open-Source Benchmark for Embedded Multiprocessor Systems
- Boomerang: Reducing Power Consumption of Response Packets in NoCs with Minimal Performance Impact
- The Accelerator Store framework for high-performance, low-power accelerator-based systems
- Centralized Adaptive Routing for NoCs
- Fractal Consistency: Architecting the Memory System to Facilitate Verification
Some interesting papers in Control Systems Magazine:
- From Control Systems to Control Software, a domain specific framework.
- Kalman Filtering with Newton’s Method [Lecture Notes]
Some interesting papers in IET Computer Vision Dec 2010:
- New multi-resolution image stitching with local and global alignment
- Automatic identification of landmarks in digital images
- Study on the performance of moments as invariant descriptors for practical face recognition systems
- Object tracking using AM-FM image features
- Nearest-neighbour ensembles in lasso feature subspaces
ParMiBench – An Open-Source Benchmark for Embedded Multiprocessor Systems, to appear in Computer Architecture Letters. The source code is available.
The Accelerator Store framework for high-performance, low-power accelerator-based systems, to appear in Computer Architecture Letters. Experiments are performed in both FPGA and ASIC.
Measuring the Gap Between FPGAs and ASICs, in TCAD 2007. In conclusion, the FPGA is 40 times larger, 3.2 times slower, and consumes 12 times more power than ASIC.
Parallel programming: can we PLEASE get it right this time?, by Tim Mattson, in DAC’08.
Parallel Programming Models Overview , by John Owens, in SIGGRAPH 2008: Beyond Programmable Shading.
More details of Nvidia Echelon research system. Simply by looking at the figures you will know who designs it. It carries the DNA of ELM, Imagine, high radix router, Dragonfly network, smart memory, etc.
A Conversation with Ed Catmull , by Pat Hanrahan.
http://corporaal.org/, the domain of Henk Corporaal.
IEEE Micro has a special issue on multicore research in Europe. The most up-to-date and comprehensive overview of European multicore research.
The Proceedings of IEEE has a special issue on Nanoelectronics Research for Beyond CMOS Information Processing.
An Analytical Approach for Network-on-Chip Performance Analysis, in TCAD 2010. The motivation of using analytical model also applies to other aspects of computer architecture research.
Some interesting papers from TPAMI Jan 2011:
An interesting article on Intel’s Stellarton CPU+FPGA chip.
The proceeding of SAMOS 2010 was released. There are some interesting papers:
- SIMD performance in software based mobile video coding, experiments performed using ARM NEON SIMD extension.
- Fast Huffman decoding by exploiting data level parallelism, experiments performed using Silicon Hive platform.
- Real-time stereo vision system using semi-global matching disparity estimation: Architecture and FPGA-implementation, impressive performance.
- Compile-time GPU memory access optimizations, by Gert-Jan van den Braak.
- OpenCL-based design methodology for application-specific processors, mapping OpenCL to Transport Triggered Architecture, from Tampere.
- A Polymorphic Register File for matrix operations, from Delft.
- Interleaving granularity on high bandwidth memory architecture for CMPs, from BSC.
- Energy-aware design space exploration of registerfile for extensible processors
- Towards scalable I/O on a many-core architecture, multi-channel DRAM interface is addressed, from University of Amsterdam.
The proceeding of International Symposium on Mixed and Augmented Reality (ISMAR) 2010 was released. There are some interesting papers:
- Positioning, tracking and mapping for outdoor augmentation
- Towards real time 3D tracking and reconstruction on a GPU using Monte Carlo simulations
- Keyframe-based modeling and tracking of multiple 3D objects
- Interactive modelling for AR applications
- An automatic parallax adjustment method for stereoscopic augmented reality systems
- Video stabilization to a global 3D frame of reference by fusing orientation sensor and image alignment data
- Time-domain augmented reality based on locally adaptive video sampling
- 3DOF tracking accuracy improvement for outdoor Augmented Reality
- Camera motion tracking in a dynamic scene
- Feature tracking and object recognition on a hand-held
- Origami recognition system using natural feature tracking
3D Human Pose Recovery from Image by Efficient Visual Feature Selection, from j.cviu 2010.