summaryrefslogtreecommitdiff
path: root/python
AgeCommit message (Collapse)AuthorFilesLines
2014-05-26caffe.Net preprocessing members belong to object, not classEvan Shelhamer1-6/+7
2014-05-23follow-up on #443 to invert k channels (instead of 3)Evan Shelhamer1-1/+2
2014-05-23Correctly invert the swapping of colour channelsJames Thewlis1-1/+2
In the 'deprocess' method, get back the image with the original channel order by inverting the original transform, rather than reversing the tuple which is incorrect.
2014-05-22fix draw_net python scriptEvan Shelhamer1-2/+2
include caffe.draw for drawing functions.
2014-05-20preprocess single inputs instead of listsEvan Shelhamer3-42/+38
For compositionality and expectations.
2014-05-20windowed detection in pythonEvan Shelhamer5-462/+262
2014-05-19squash infuriating loop assignment bug in batchingEvan Shelhamer1-1/+1
2014-05-19image classification in pythonEvan Shelhamer5-129/+206
2014-05-19fix padding for the last batchEvan Shelhamer1-3/+4
2014-05-19split drawnet into module code and scriptEvan Shelhamer2-15/+25
Don't run scripts in the module dir to avoid import collisions between io and caffe.io.
2014-05-19add caffe.io submodule for conversions, image loading and resizingEvan Shelhamer3-12/+89
2014-05-19fix python mean subtractionEvan Shelhamer1-7/+9
2014-05-16drop cute names in favor of Net.{pre,de}process() for input formattingEvan Shelhamer1-44/+45
...and refer to inputs as inputs and not images since general vectors and matrices are perfectly fine.
2014-05-16Net.caffeinate() and Net.decaffeinate() format/unformat listsEvan Shelhamer1-41/+46
2014-05-16take blob args as ndarrays and assign on the python sideEvan Shelhamer2-184/+72
Take blob args and give blob returns as single ndarrays instead of lists of arrays. Assign the net blobs and diffs as needed on the python side, which reduces copies and simplifies the C++ side of the wrapper. Thanks @longjon for the suggestion.
2014-05-15resize to input dimensions when formatting in pythonEvan Shelhamer1-0/+7
2014-05-15replace iterator with indices for consistencyEvan Shelhamer1-6/+6
2014-05-15python styleEvan Shelhamer2-254/+249
2014-05-15fix accidental revert of Init() from f5c28581Evan Shelhamer1-2/+1
2014-05-14batch inputs in python by forward_all() and forward_backward_all()Evan Shelhamer2-2/+102
2014-05-14don't squeeze blob arrays for pythonEvan Shelhamer1-2/+2
Preserve the non-batch dimensions of blob arrays, even for singletons. The forward() and backward() helpers take lists of ndarrays instead of a single ndarray per blob, and lists of ndarrays are likewise returned. Note that for output the blob array could actually be returned as a single ndarray instead of a list.
2014-05-14python forward() and backward() extract any blobs and diffsEvan Shelhamer1-18/+32
2014-05-14python Net.backward() helper and Net.BackwardPrefilled()Evan Shelhamer2-2/+40
2014-05-14bad forward/backward inputs throw exceptions instead of crashing pythonEvan Shelhamer1-13/+14
2014-05-14pycaffe Net.forward() helperEvan Shelhamer1-0/+36
Do forward pass by prefilled or packaging input + output blobs and returning a {output blob name: output list} dict.
2014-05-14set input preprocessing per blob in pythonEvan Shelhamer1-22/+64
2014-05-14expose input and output blob names to python as listsEvan Shelhamer1-0/+20
2014-05-13pycaffe comments, lintEvan Shelhamer1-11/+11
2014-05-13add python io getters, mean helper, and image caffeinator/decaffeinatorEvan Shelhamer1-0/+73
2014-05-13make python wrapper mean match binaryproto dimensionsEvan Shelhamer1-0/+0
ilsvrc_2012_mean.npy has dims K x H x W. Code written for the old D x D x K mean needs to be rewritten!
2014-05-13match existing python formattingEvan Shelhamer1-16/+21
2014-05-02pycaffe: allow 1d labels to be passed to set_input_arraysJonathan L Long2-13/+24
2014-05-02pycaffe: add Net.set_input_arrays for input from numpyJonathan L Long1-1/+65
This requires a net whose first layer is a MemoryDataLayer.
2014-05-02pycaffe: store a shared_ptr<CaffeNet> in SGDSolverJonathan L Long1-1/+5
Doing this, rather than constructing the CaffeNet wrapper every time, will allow the wrapper to hold references that last at least as long as SGDSolver (which will be necessary to ensure that data used by MemoryDataLayer doesn't get freed).
2014-05-02pycaffe: let boost pass shared_ptr<CaffeNet>Jonathan L Long1-1/+1
2014-04-25pycaffe: add unary CaffeNet constructor for uninitialized netsJonathan L Long1-3/+15
Note that parameters are uninitialized, not zero-filled.
2014-04-25note pydot dependency of model visualizationEvan Shelhamer1-1/+6
2014-04-25Update the drawnet.py to reflect the recent revised net definition.ZhiHeng NIU1-4/+11
2014-04-11fix lint errors by adding 'explicit' to new single arg pycaffeJeff Donahue1-2/+2
constructors
2014-04-04pycaffe: expose SGDSolver.solveJonathan L Long1-1/+8
2014-04-04pycaffe: introduce CheckFile helperJonathan L Long1-23/+15
2014-04-04pycaffe: expose SGDSolver.netJonathan L Long1-1/+7
2014-04-04switch from inheritance to directly overriding methods for caffe.NetJonathan L Long2-22/+28
Also, internally drop the "Caffe" prefix for classes exported by the internal _caffe module. This change means that SGDSolver will automatically return a Net with the "nice" Python interface rather than the internal one, (and avoids the extra layer of method wrapping that results from switching to composition).
2014-04-04pycaffe: expose SGDSolverJonathan L Long3-2/+23
2014-04-04Revert "Fix segfault." (python)Evan Shelhamer1-1/+0
Revert 88507032de1eb675539d91a1d79c221f3faf5d99 from #287. The call to `Py_Initialize()` is unnecessary and masked a build configuration issue.
2014-04-04Merge pull request #277 from jeffdonahue/makefile-improvementsEvan Shelhamer1-0/+0
Makefile improvements
2014-04-02Fix segfault.Rob Hess1-0/+1
2014-04-02cleanup python build and cleanJeff Donahue1-0/+0
2014-03-27Standardize copyright, add root-level CONTRIBUTORS creditEvan Shelhamer1-1/+1
2014-03-18drop models/ in favor of examples/Evan Shelhamer1-2/+2
Move the Caffe reference imagenet model and script to fetch it to examples/imagenet. Caffe bundles reference models, but it makes more sense to keep them close to examples.
generated by cgit v1.2.3 (git 2.25.1) at 2024-06-09 18:36:08 +0000