Home | History | Annotate | Download | only in cuda

Lines Matching refs:trainIdx

127                 const int trainIdx = t * BLOCK_SIZE + threadIdx.x;
129 if (queryIdx < query.rows && trainIdx < train.rows && distVal < bestDistance && mask(queryIdx, trainIdx))
133 bestTrainIdx = trainIdx;
171 const PtrStepSzi& trainIdx, const PtrStepSzf& distance,
179 matchUnrolledCached<BLOCK_SIZE, MAX_DESC_LEN, Dist><<<grid, block, smemSize, stream>>>(query, train, mask, trainIdx.data, distance.data);
230 const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance,
238 matchUnrolledCached<BLOCK_SIZE, MAX_DESC_LEN, Dist><<<grid, block, smemSize, stream>>>(query, trains, n, mask, trainIdx.data, imgIdx.data, distance.data);
287 const int trainIdx = t * BLOCK_SIZE + threadIdx.x;
289 if (queryIdx < query.rows && trainIdx < train.rows && distVal < bestDistance && mask(queryIdx, trainIdx))
293 bestTrainIdx = trainIdx;
329 const PtrStepSzi& trainIdx, const PtrStepSzf& distance,
337 matchUnrolled<BLOCK_SIZE, MAX_DESC_LEN, Dist><<<grid, block, smemSize, stream>>>(query, train, mask, trainIdx.data, distance.data);
386 const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance,
394 matchUnrolled<BLOCK_SIZE, MAX_DESC_LEN, Dist><<<grid, block, smemSize, stream>>>(query, trains, n, mask, trainIdx.data, imgIdx.data, distance.data);
442 const int trainIdx = t * BLOCK_SIZE + threadIdx.x;
444 if (queryIdx < query.rows && trainIdx < train.rows && distVal < bestDistance && mask(queryIdx, trainIdx))
448 bestTrainIdx = trainIdx;
484 const PtrStepSzi& trainIdx, const PtrStepSzf& distance,
492 match<BLOCK_SIZE, Dist><<<grid, block, smemSize, stream>>>(query, train, mask, trainIdx.data, distance.data);
540 const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance,
548 match<BLOCK_SIZE, Dist><<<grid, block, smemSize, stream>>>(query, trains, n, mask, trainIdx.data, imgIdx.data, distance.data);
560 const PtrStepSzi& trainIdx, const PtrStepSzf& distance,
565 matchUnrolledCached<16, 64, Dist>(query, train, mask, trainIdx, distance, stream);
569 matchUnrolledCached<16, 128, Dist>(query, train, mask, trainIdx, distance, stream);
573 matchUnrolled<16, 256, Dist>(query, train, mask, trainIdx, distance, stream);
577 matchUnrolled<16, 512, Dist>(query, train, mask, trainIdx, distance, stream);
581 matchUnrolled<16, 1024, Dist>(query, train, mask, trainIdx, distance, stream);
585 match<16, Dist>(query, train, mask, trainIdx, distance, stream);
591 const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance,
596 matchUnrolledCached<16, 64, Dist>(query, trains, n, mask, trainIdx, imgIdx, distance, stream);
600 matchUnrolledCached<16, 128, Dist>(query, trains, n, mask, trainIdx, imgIdx, distance, stream);
604 matchUnrolled<16, 256, Dist>(query, trains, n, mask, trainIdx, imgIdx, distance, stream);
608 matchUnrolled<16, 512, Dist>(query, trains, n, mask, trainIdx, imgIdx, distance, stream);
612 matchUnrolled<16, 1024, Dist>(query, trains, n, mask, trainIdx, imgIdx, distance, stream);
616 match<16, Dist>(query, trains, n, mask, trainIdx, imgIdx, distance, stream);
624 const PtrStepSzi& trainIdx, const PtrStepSzf& distance,
630 trainIdx, distance,
636 trainIdx, distance,
641 template void matchL1_gpu<uchar >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
642 //template void matchL1_gpu<schar >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
643 template void matchL1_gpu<ushort>(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
644 template void matchL1_gpu<short >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
645 template void matchL1_gpu<int >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
646 template void matchL1_gpu<float >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
649 const PtrStepSzi& trainIdx, const PtrStepSzf& distance,
655 trainIdx, distance,
661 trainIdx, distance,
666 //template void matchL2_gpu<uchar >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
667 //template void matchL2_gpu<schar >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
668 //template void matchL2_gpu<ushort>(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
669 //template void matchL2_gpu<short >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
670 //template void matchL2_gpu<int >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
671 template void matchL2_gpu<float >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
674 const PtrStepSzi& trainIdx, const PtrStepSzf& distance,
680 trainIdx, distance,
686 trainIdx, distance,
691 trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
692 //template void matchHamming_gpu<schar >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
693 template void matchHamming_gpu<ushort>(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
694 //template void matchHamming_gpu<short >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
695 template void matchHamming_gpu<int >(const PtrStepSzb& queryDescs, const PtrStepSzb& trainDescs, const PtrStepSzb& mask, const PtrStepSzi& trainIdx, const PtrStepSzf& distance, cudaStream_t stream);
698 const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance,
704 trainIdx, imgIdx, distance,
710 trainIdx, imgIdx, distance,
715 template void matchL1_gpu<uchar >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
716 //template void matchL1_gpu<schar >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
717 template void matchL1_gpu<ushort>(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
718 template void matchL1_gpu<short >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
719 template void matchL1_gpu<int >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
720 template void matchL1_gpu<float >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
723 const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance,
729 trainIdx, imgIdx, distance,
735 trainIdx, imgIdx, distance,
740 //template void matchL2_gpu<uchar >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
741 //template void matchL2_gpu<schar >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
742 //template void matchL2_gpu<ushort>(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
743 //template void matchL2_gpu<short >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
744 //template void matchL2_gpu<int >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
745 template void matchL2_gpu<float >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& maskCollection, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
748 const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance,
754 trainIdx, imgIdx, distance,
760 trainIdx, imgIdx, distance,
765 template void matchHamming_gpu<uchar >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
766 //template void matchHamming_gpu<schar >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
767 template void matchHamming_gpu<ushort>(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
768 //template void matchHamming_gpu<short >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);
769 template void matchHamming_gpu<int >(const PtrStepSzb& query, const PtrStepSzb& trains, const PtrStepSz<PtrStepb>& masks, const PtrStepSzi& trainIdx, const PtrStepSzi& imgIdx, const PtrStepSzf& distance, cudaStream_t stream);