Skip to content

Instantly share code, notes, and snippets.

View crazysal's full-sized avatar
🎯
Focusing

Saleem Ahmed crazysal

🎯
Focusing
View GitHub Profile
@crazysal
crazysal / linking with C C++ libraries with gcc g++ when building an existing project
Created October 15, 2020 11:30
linking with C/C++ libraries with gcc/g++ when building an existing project
Notes on linking with C/C++ libraries with gcc/g++ when building an
existing project.
-- Iain Murray, 2015.
There are at least three things that can go wrong when trying to link
with a library that is in a non-standard location on your machine:
1. The compiler can't find the .h header files.
2. The linker can't find the library's binary .a or .so files.
@crazysal
crazysal / cuda_ver.txt
Created April 25, 2019 18:19
Check Cuda and cuDNN versions
cat /usr/local/cuda/version.txt
nvcc --version
cat /usr/local/cuda/include/cudnn.h | grep CUDNN_MAJOR -A 2
or
cat /usr/include/cudnn.h | grep CUDNN_MAJOR -A 2
@crazysal
crazysal / accmathVOCtriplet.py
Created March 5, 2019 20:05
Access Math Voc Triplet Data Loader
import os
import traceback
import time
import xml.etree.ElementTree as ET
import torch
from torch.utils.data import Dataset
from data.utils import filter_by_shape, extend_bbox, image_channel_mean, bbox_to_quad
from data.transforms import TransformCrop, TransformPad, BinarizerWeightGeneration
@crazysal
crazysal / lstm.pt
Created January 10, 2019 18:14
simple LSTM PT in caffe
name: "LSTM"
input: "data"
input_shape { dim: 320 dim: 1 }
input: "clip"
input_shape { dim: 320 dim: 1 }
input: "label"
input_shape { dim: 320 dim: 1 }
layer {
name: "Silence"
type: "Silence"
@crazysal
crazysal / Decoder.pt
Last active January 10, 2019 17:21
Decoder training model for end to end textspotter
name : "DECODER"
input: "sample_gt_label_input"
input_shape { dim: 1 dim: 1 }
input2: "sample_gt_cont"
input_shape { dim: 1 dim: 1 }
input: "decoder"
input_shape { dim: 1 dim: 1 dim: 256 }
input: "sample_gt_cont"
input_shape { dim: 1 dim: 1 }
input: "sample_gt_label_input"
input_shape { dim: 1 dim: 1 }
layer {
name: "PVANET"
input: "data"
input_shape { dim: 1 dim: 3 dim: 512 dim: 512 }
input: "fcn_th"
input_shape { dim: 1 dim: 1 }
################################################################################
## Convolution
name: "PVANET"
layer {
type: "YOU DATA LAYER"
top: "mask_gt" ### 2 * 640 * 640 * 1 ### mask for text/non-text segmentation
top: "mask_iou_angle" ### 1 * 640 * 640 * 5 ### iou with 4 channels and angle with one channel
top: "gt_bbox" ### N * 8 ### grounding truth boxes for text (for computing loss)
top: "ignore_bbox" ### N * 8 ### grounding truth boxes for text (for ignoring)
}
@crazysal
crazysal / cropAndResize.c
Last active December 30, 2018 00:27
C code to crop and resize a rectangle box from an image to given dimension
void CropAndResizePerBox(
const float * image_data,
const int batch_size,
const int depth,
const int image_height,
const int image_width,
const float * boxes_data,
const int * box_index_data,
const int start_box,
@crazysal
crazysal / aes_example.c
Created September 11, 2018 20:17
AES Example code by intel sample library
/* ex: set ts=4: */
/*
based on code from http://www.progressive-coding.com/tutorial.php?id=3#end
---------------------------------------------------------------------------
Copyright (c) 2007-2010, Laurent Haan, Luxembourg, LU. All rights reserved.
LICENSE TERMS
The redistribution and use of this software (with or without changes)
is allowed without the payment of fees or royalties provided that: